ML23104A054
| ML23104A054 | |
| Person / Time | |
|---|---|
| Site: | 05000610 |
| Issue date: | 08/15/1996 |
| From: | Peretz F Lockheed Martin Energy Systems |
| To: | Office of Nuclear Material Safety and Safeguards, US Dept of Energy, Office of Environmental Management |
| References | |
| DE-AC05-84OR21400 ORNL/ER-380 | |
| Download: ML23104A054 (1) | |
Text
{{#Wiki_filter:~
&OCKHEEO MARFIN:,4--
ORNL/ER-380 ENVIRONMENTAL RESTORATION Identification and Evaluation PROGRAM of Alternatives for the Disposition of Fluoride Fuel and Flush Salts from the Molten Salt Reactor Experiment at Oak Ridge National Laboratory, Oak Ridge, Tennessee This document has been approved by the ORNL Technical lnfomrntion Oicelqlt, for release to the public. Date: i IS' ENERGYSYSTEMS MANAGEDBY LOCKHEED MARTIN ENERGY SYSTEMS, INC. FOR THE UNITED STATES DEPARTMENT OF ENERGY UCN-17560 (6 8-95) ER
Jacobs Engineering Group, Inc., and Advanced Integrated Management Services contributed to the preparation of this document and should not be considered eligible contractors for its review. This report has been reproduced directly from the best available copy. Available to DOE and DOE contractors from the Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831; prices available from 423-576-8401 (fax 423-576-2865). Available to the public from the National Technical Information Service, U.S. Department of Commerce, 5285 Port Royal Rd., Springfield, VA 22161.
DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsi-bility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Refer- ORNL/ER-380 ence herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom-mendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. Energy Systems Environmental Restoration Program Identification and Evaluation of Alternatives for the Disposition of Fluoride Fuel and Flush Salts from the Molten Salt Reactor Experiment at Oak Ridge National Laboratory, Oak Ridge, Tennessee F. J. Peretz Date Issued-August 1996 Prepared for the U.S. Department of Energy Office of Environmental Management under budget and reporting code EW 20 Environmental Management Activities at the OAK RIDGE NATIONAL LABORATORY Oak Ridge, Tennessee 37831-6285 managed by LOCKHEED MARTIN ENERGY SYSTEMS, INC. for the U.S. DEPARTMENT OF ENERGY under contract DE-AC05-84OR21400 DISTRIBUTION OF THIS DOCUMENT IS UNUMITE~
DISCLAIMER Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.
PREFACE This report on the Identification and Evaluation ofAlternatives for lhe Disposition ofFluoride Fuel and Flush Salts from /he Molten Sall Reac/or Experimenl al Oak Ridge National Laboralory, Oak Ridge, Tennessee, (ORNL/ER-380) was prepared as part of the Decontamination and Decommissioning activities taking place at Oak Ridge National Laboratory under the auspices of the Environmental Restoration Program. It was prepared under Work Breakdown Structure 1.4.12.6.2.01.05.06, Activity Data Sheet 3701, .. ORNL Decontamination and Decommissioning Program" in support of the CERCLA feasibility study being prepared by Jacobs Engineering Group, Inc. iii
ACKNOWLEDGMENTS The work reported herein is a compilation of the efforts of DOE and contractor personnel from the Environmental Restoration Program, Lockheed Martin Energy Systems Central Engineering Services, Oak Ridge National Laboratory, Jacobs Engineering, Advanced Integrated Management Services, Argonne National Laboratory, Idaho National Engineering Laboratory, the Savannah River Site, and other DOE and commercial contacts. References to sources of specific information are provided whenever practical. V
CONTENTS PREFACE ................. ................. ................. ................. .. iii ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V FIGURES ................. ................. ................. ................. ... xi TABLES ................. ................. ................. ................. .. xiii ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv EXECUTIVE
SUMMARY
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
- 1. OVERVIEW ................. ................. ................. ............. 1-1 1.1 OVERVIEW OF THE MOLTEN SALT REACTOR EXPERIMENT FUEL SALT DISPOSITION TASK ................. ................. ................. . 1-1 LI.I Facility Description ................. ................. .............. 1-1 1.1.2 Operating History ................. ................. ............... 1-6 1.1.3 Postoperation History ................. ................. ............ 1-11 1.1.4 MSRE Remediation Project ................. ................. ....... 1-15 1.1.5 Fuel and Flush Salt Disposition Task ................. ................ 1-18 1.2 CHARACTERIZATION OF THE MSRE FUEL AND FLUSH SALTS ............ 1-19 1.2. l Inventory and Composition of the Fuel and Flush Sall ................. ... 1-19 1.2.2 Fissile Material Inventory ................. ................. ......... 1-20 1.2.3 Radionuclide Inventory ................. ................. ............ 1-22 1.2.4 Chemical Characteristics ................. ................. ......... 1-26 1.2.5 Summary of All Locations Containing Fuel or Flush Salt ................. . 1-31 1.3 PRESENT CONDITION OF THE MSRE FACILITY ................. ......... 1-33 1.4 BASELINE RISK ELEMENTS ................. ................. .......... 1-36 1.5 OUTLINE OF THE ALTERNATIVES EVALUATION PROCESS ............... 1-38
- 2. POTENTIAL END POINTS ................. ................. ................. 2-1 2.1 MSRE DRAIN TANK CELL ................. ................. ............. 2-1 2.2 REUSE ................. ........ *................. ................. ..... : 2-1 2.3 WASTE ISOLATION PILOT PLANT ................. ................. ...... 2-2 2.3. l General Description ................. ................. .............. 2-2 2.3.2 Regulatory Basis ................. ................. ................. 2-8 2.3.3 Disposal of MSRE Salts in WIPP ................. ................. ... 2-9 2.4 FEDERAL REPOSITORY ................. ................. .............. 2-16 2.4.1 General Description ................. ................. ............. 2-16 2.4.2 Regulatory Basis ................. ................. ................ 2-17 2.4.3 Disposal of MSRE Material in the Federal Repository ................. ... 2-19 2.5 FISSILE MATERIALS DISPOSITION PROGRAM ................. .......... 2-20 2.6 LOW-LEVEL WASTE STORAGE FACILITIES ................. ............. 2-23 2.7 INTERIM STORAGE ................. ................. ................. . 2-24 vii
- 3. IBCI:IN"OLOGIES .......... .......... .......... .......... .......... ...... 3-1 3.1 REMOVAL IBCI:IN"OLOGIES .......... .......... .......... .......... .. 3-1 3.1.1 Features of the Drain Tanks .......... .......... .......... ......... 3-1 3.1.2 Removal as Molten Salt .......... .......... .......... .......... .. 3-1 3.1.3 Removal as Solid Salt .......... .......... .......... .......... .. 3-10 3.1.4. Removal with a Liquid Solvent/Carrier .......... .......... ......... 3-17 3.1.5 Screening of Removal Technologies .......... .......... .......... . 3-20 3.2 SEPARATION IBCI:IN"OLOGIES .......... .......... .......... ........ 3-21 3.2.1. Fluoride Volatility .......... .......... .......... .......... ..... 3-21 3.2.2. Electrometallurgical Separation .......... .......... .......... .... 3-24 3.2.3. Vacuum Distillation of Salt .......... .......... .......... ........ 3-32 3.2.4. Other Uranium Separation Technologies .......... .......... ....... 3-34 3.2.5. Screening of Processing Technologies .......... .......... ......... 3-36 3.3 STABILIZATION IBCI:IN"OLOGIES .......... .......... .......... ...... 3-37 3.3.1. Stabilization of Fluoride Salt with a Getter .......... .......... ...... 3-37 3.3.2. Conversion of Fluoride Salt to Oxide Using the INEL New Waste Calcining Facility .......... .......... .......... .......... ...... 3-38 3.3.3. Conversion of Fluoride Salt to Borosilicate Glass .......... .......... 3-41 3.3.4 Other Salt Conversion Processes .......... .......... .......... .... 3-48 3.3.5 Conversion of Uranium to U3O8 ********** ********** ********** **** 3-51 3.3.6 Screening of Stabilization Technologies .......... .......... ........ 3-53 3.4 PACKAGIN G AND TRANSPORTATION .......... .......... .......... .. 3-54 3.4.1. Salt Packaging and Transportation .......... .......... .......... .. 3-54 3.4.2. Uranium Packaging .......... .......... .......... .......... .... 3-55 3.4.3. Packaging and Transportation of Other Waste Forms .......... ........ 3-55 3.4.4. Licensing, Regulatory, and Other Issues .......... .......... ........ 3-55
- 4. IDENTIFICATION OF ALIBRNAT IVES .......... .......... .......... ...... 4-1 4.1 PERMANE NT DISPOSAL IN THE DRAIN TANKS .......... .......... .... 4-1 4.2 DISPOSAL OF ALL KEY CONTAMINANTS IN THE FEDERAL REPOSITORY 4-2 4.3 DISPOSAL OF SEPARAIB D URANIUM .......... .......... .......... ... 4-5 4.4 DISPOSAL OF THE KEY CONTAMINANTS IN THE SALT RESIDUE IN WIPP .......... .......... .......... .......... ..... 4-5 4.5 DISPOSAL OF KEY CONTAMINANTS IN SALT RESIDUE IN THE FEDERAL REPOSITO RY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7 4.6 REUSE OF THE SALT .......... .....*.... .......... .......... ........ 4-8 4.7 INIBRIM STORAGE .......... .......... .......... .......... ........ 4-12
- 5. EVALUAT ION OF ALIBRNAT IVES .......... .......... .......... ......... 5-1 5.1 EVALUAT ION MATRIX .......... .......... .......... .......... ...... 5-1 5.2 SCORINGR ATIONAL EFOREAC HCRITERI ON .......... .......... ..... 5-1 5 .2.1 Overall Protection of Human Health and the Environment . . . . . . . . . . . . . . 5-1 5.2.2 Compliance with Applicable or Relevant and Appropriate Requirements . . . 5-1 5.2.3 Long-Term Effectiveness and Permanence .......... .......... ....... 5-1 5.2.4 Reduction of Toxicity, Mobility, or Volume Through Treatment ......... 5-3 5.2.5 Short-Term Effectiveness .......... .......... .......... .......... 5-3 5.2.6 Implementability .......... .......... .......... .......... ....... 5-3 5.2.7 Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4 viii
5.2.8 Adaptability to Changing Requirements ....... ....... ....... ....... .... 5-5 5.2.9 Compatibility With Programmatic Objectives ........ ........ ........ .... 5-5 5.3 RECOMMENDATIONS ....... ....... ....... ....... ....... ....... ....... .. 5-6 5.4 REVIEW OF THE ALTERNATE APPROACHES TO INTERIM STORAGE ....... 5-7
- 6. IDENTIFICATION OF KEY UNKNOWNS, POTENTIAL CONSEQUENCES, AND PATH TO RESOLUTION ....... ....... ....... ....... ....... ....... ...... 6-1 6.1 CERCLA PROCESS ....... ....... ....... ....... ....... ....... ....... ..... 6-1 6.2 WASTE CLASSIFICATION ....... ....... ....... ....... ....... ....... .....
6-1 6.3 SALT CHEMISTRY ....... ....... ....... ....... ....... ....... ....... ..... 6-2 6.4 AVAILABILITY OF SUPPORTING FACILITIES ........ ........ ........ ..... 6-2 6.5 TRANSPORTATION AND INTERSTATE AGREEMENTS ........ ........ ..... 6-2 6.6 FUNDING ....... ....... ....... ....... ....... ....... ....... ....... ...... 6-3
- 7. REFERENCES ....... ....... ....... ....... ....... ....... ....... ....... ...... 7-1 ix
FIGURES 1.1 Arrangement of the principal components of the MSRE ............ ............ . 1-2 1.2 Plan view of the MSRE facility ............ ............ ............ ........ 1-3 1.3 Sketch of an MSRE fuel drain tank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4 1.4 Cross-section through the MSRE drain tank cell and adjacent areas ............ .... 1-5 1.5 Flowsheet of the MSRE salt processing system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-7 1.6 Shield block arrangement on top of the drain tank cell ............ ............ .. 1-8 I. 7 Typical plan view of setup of the maintenance shield . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-9 1.8 Arrangement of the equipment used to add 233 U directly into an MSRE fuel drain tank 1-1 O 1.9 Principal components of the MSRE off-gas system ............ ............ .... 1-13 1.10 Location of the uranium deposit in the auxiliary charcoal bed ............ ....... 1-14
- 1. 11 Flowsheet of the reactive gas removal system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-17 1.12 Decay chains for 232U and 233U ............ ............ ............ ........ 1-24 1.13 Fission product activity ofMSRE fuel and flush salts ............ ............ .. 1-27 1.14 Actinide and daughter activity ofMSRE fuel and flush salts ............ ........ 1-28 1.15 Phase diagram of the LiF-BeF2-ZrF4 system ............ ............ ......... 1-29 1.16 Projected fluorine generation resulting from fission product decay ............ ... 1-32 1.17 Approximate elevation to which water has risen in sump room with sump pumps out of operation ............ ............ ............ ............ ........ 1-37 1.18 Logic diagram of the alternatives identification process ............ ............ 1-39 1.19 Expanded logic of the disposal evaluation block ............ ............ ...... 1-40 2.1 Location of the Waste Isolation Pilot Plant, indicating transportation routes from major DOE sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 2.2 Geologic formations at the Waste Isolation Pilot Plant ............ ............ .. 2-4 2.3 Arrangement of the remote handled transuranic waste disposal facilities at WIPP ..... 2-6 2.4 Dimensions of a remote handled transuranic waste canister . . . . . . . . . . . . . . . . . . . . . . 2-7 2.5 Proposed shielded container for MSRE salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-10 2.6 Depiction of three MSRE salt packages in an RH-TRU waste canister ............ . 2-11 2.7 Depiction of the federal waste repository under consideration at the Yucca Mountain site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-18 2.8 Site plan for the Radiochemical Development Facility (Building 3019) at ORNL .... 2-21 2.9 Storage wells for 233 U between cells II and III of the ORNL Radiochemical Development Facility ............ .... *. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-22 3.1 Components of the fuel and flush tank system ............ ............ ......... 3-2 3.2 Arrangement of the fuel and flush tanks in the drain tank cell, including cell shielding blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 3.3 Reduction potentials for metals ofinterest in the MSRE drain tanks ............ ... 3-5 3 .4 Diagram of process to melt a pool salt with an internal heater while restoring the fluorine balance with a HF/H2 sparge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-9 3.5 Depiction of the hardware needed to mechanically remove the salt using a CO2 blaster mounted to an overhead manipulator system . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-11 3 .6 Depiction of the hardware needed to mechanically remove the salt using a CO2 blaster mounted directly onto the drain tank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12 xi
3.7 Depiction of the wand and nozzle for the CO2 ablation process . . . . . . . . . . . . . . . . . . 3-13 3.8 Depiction of a system to mechanically remove the salt from the tanks using a vacuum auger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-15 3 .9 Depiction of a system to withdraw the cooling thimbles from a fuel salt drain tank . . . 3-16 3 .10 Depiction of the general arrangement of equipment and shielding needed to remove a fuel drain tank and salt as a unit and transport it to another location . . . . . . . 3-18 3.11 Depiction of a cask system that could be used to remove a fuel drain tank and salt as a unit and transport it to another location . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-19 3.12 Flowsheet of the MSRE salt processing system ........ ........ ........ ....... 3-22 3.13 Overall dimensions for a shielded inert atmosphere enclosure to house the Argonne electrorefiner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-25 3.14 Diagram of the processing steps used to electrometallurgically separate radioactive components from the fuel salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-26 3.15 Depiction of the crucible handling activities for the Argonne electrorefiner ........ 3-29 3 .16 Depiction of the solid cathode handling activities for the Argonne electrorefiner . . . . 3-30 3 .17 Depiction of the bismuth electrode preparation and salt cleaning from the pounder cathode steps activities for the Argonne electrorefiner . . . . . . . . . . . . . . . . . . 3-31 3.18 Simplified flow diagram of the MSRE distillation experiment ........ ........ ... 3-33 3.19 Diagram of the overall operation of the INEL New Waste Calcining Facility ....... 3-39 3.20 A typical borosilicate glass high level waste package ........ ........ ........ .. 3-42 3.21 Diagram of the overall operations performed at the Savannah River Defense Waste Processing Facility ........ ........ ........ ........ ........ ........ 3-43 3.22 Diagram of the overall operations that would be performed at the proposed INEL waste immobilization facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-46 3.23 Diagram of the overall operations that would be performed at the proposed INEL vitrification facility ........ ........ ........ ........ ........ ........ 3-47 3.24 Process steps for batch processing using the glass materials oxidation and dissolution system ........ ........ ........ ........ ........ ........ ...... 3-49 3.25 Proposed overall flowsheet for the conversion ofUF to oxide ........ ........ ... 3-52 6 4.1 Flowcharts for alternative 1, no action, and alternative 2, enhanced storage ......... . 4-3 4.2 Flowchart for alternative 3, dispose of the salt, including uranium, in the federal repository . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4 4.3 Flowchart for alternative 4, transfer uranium to the materials disposition program and dispose of the salt residue in WIPP ........ ........ ........ ........ ...... 4-6 4.4 Flowchart for alternative Sa, transfer uranium to tlie materials disposition program and dispose of the salt residue in the federal repository, based on vitrification at the DWPF 4-9 4.5 Flowchart for alternative Sb, transfer uranium to the materials disposition program and dispose of the salt residue in the federal repository, based on electrorefining molten fluoride salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-10 4.6 Flowchart for alternative 6, tr~sfer uranium to the materials disposition program and transfer the salt to another program for reuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-11 4.7 Flowchart for alternative 7, transfer uranium to the materials disposition program and place the salt residue in interim storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-14 xii
TABLES 1.1 Results from analysis of Molten Sall Reactor Experiment off-gas samples taken in 1994 . 1-12 1.2 Inventory of salt stored in the fuel and flush drain tanks ......... ......... ......... . 1-21 1.3 Composition of the fuel and flush salts ......... ......... ......... ......... ..... 1-21 1.4 Mass of uranium and plutoniwn in fuel and flush salt, kg ......... ......... ........ 1-22 1.5 Some key criticalily safely parameters for fissionable isotopes present in the MSRE salt ......... ......... ......... ......... ......... ......... .... 1-22 1.6 Inventory of fission and activation product isotopes ......... ......... ......... .... 1-23 I. 7 Inventory of actii;iide isotopes ......... ......... ......... ......... ......... ... 1-25 1.8 Dose at 1 m for unshielded radionuclides in postulated fuel and flush salt packages ...... 1-26 1.9 Crystallization sequence for Molten Salt Reactor Experiment fuel salt ......... ....... 1-30 1.10 Summary of Molten Salt Reactor Experiment salt inventory ......... ......... ...... 1-33 2.1 Radionuclide inventory in all of the fuel and flush salts sent to the repository (Ci) ...... 2-12 2.2 Dose rates at side surface of postulated fuel and flush salt containers ......... ........ 2-12 2.3 Comparison of fuel and flush sail parameters in a remote-handled transuranic canister to Waste Isolation Pilot Plant (WIPP) waste acceptance criteria ......... ......... ..... 2-13 2.4 Definitions of waste classifications ......... ......... ......... ......... ........ 2-14 3.1 Solubility of components in MSRE fuel salt in various solvents (g/L at 25 ° C) ......... 3-17 3.2 Process steps and products for electrochemical trealment of MSRE fuel salt ......... .. 3-27 5.1 Alternatives evaluation matrix for the Mallen Salt Reaclor Experiment Fuel Salt Disposition task ......... ......... ......... ......... ......... ......... ...... 5-2 5.2 Evaluation matrix for alternate approaches to alternative 7, transfer uranium to the materials distribution program and pince the salt residues in interim storage ......... 5-9
ABBREVIATIONS ANL-W Argonne National Laboratory-West CERCLA Comprehensive Environmental Response, Compensation, and Liability Act CEUSP Consolidated Edison Uranium Solidification Project CFR Code ofFederal Regulations CH-TRU contact-handled transuranic DNFSB Defense Nuclear Facilities Safety Board DOE U.S. Department of Energy DWPF Defense Waste Processing Facility EIS environmental impact statement EPA U.S. Environmental Protection Agency FFA Federal Facilities Agreement GMODS Glass Material Oxidation and Dissolution System HEP A high-efficiency particulate air ICPP Idaho Chemical Processing Plant INEL Idaho National Engineering Laboratory MSBR Molten Salt Breeder Reactor MSRE Molten Salt Reactor Experiment MTR Materials Test Reactor NRC Nuclear Regulatory Commission ORNL Oak Ridge National Laboratory RCRA Resource Conservation and Recovery Act RDF Radiochemical Development Facility RH-TRU remote-handled transuranic SRS Savannah River Site TDEC Tennessee Department ofEnvironn1ent and Conservation WIPP Waste Isolation Pilot Plant WTWBIR WIPP Transuranic Waste Baseline Inventory Report xv
EXECUTIVE
SUMMARY
This document presents an initial identification and evaluation of the alternatives for disposition of the fluoride fuel and flush salts stored in the drain tanks at the Molten Salt Reactor Experiment (MSRE) at Oak Ridge National Laboratory (ORNL). It will serve as a resource for the U.S. Department of Energy contractor preparing the feasibility study for this activity under the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA). This document will also facilitate further discussion on the range of credible alternatives, and the relative merits of alternatives, throughout the time that a final alternative is selected under the CERCLA process. On the basis of the review prepared under this study, the following sequence of preferences was defined: If reuse of the fuel and/or flush salts is a serious option, then it should be selected. If any of the geologic disposal alternatives prove technically and programmatically implementable, every reasonable effort should be made to implement that alternative. In particular, efforts should be made to resolve the regulatory and programmatic obstacles to disposal in the Waste Isolation Pilot Plant (WIPP). At the present time, only the interim storage option is likely to be implementable. It does provide adequate protection of human health and the environment, but at some time an ultimate disposition must be identified to end the cost of perpetual care and monitoring. Storage of fluorinated, stabilized salt can meet the quantitative requirements for disposal at WIPP or can serve as the feed for any of the other processes evaluated. Thus, interim storage is recommended as being consistent with ultimate disposition and being fully implementable at this time. Pennanent disposal of the salts in their drain tanks is not recommended. The long-lived actinides in the salt should be placed in geologic disposal; the penetrating radiation from fission products and uranium decay daughters will require that the waste be remote-handled. Specific scenarios for handling and disposition of the salt include a potential for the reusing the salts at Los Alamos National Laboratory; separating the uranium and chemically stabilizing the salt; calcining and possibly eventually vitrifying the salt at Idaho National Engineering Laboratory; blending the salt into the feed stream to the Defense Waste Processing Facility at the Savannah River Site; and constructing an electrorefining system or other new process at ORNL or at another site. Interim storage opportunities exist at any of the above sites. Chapter I of this report presents backgrmmd information on MSRE, the salts, and the methodology used to define and evaluate alternatives for the disposition of the MSRE fuel and flush salts. Chapter 2 identifies potential ultimate end points for the key contaminants in the salts. Chapter 3 summarizes the technologies available for removal, processing, and stabilization of the salt. The disposition alternatives are identified in Chap. 4 and evaluated in Chap. 5. Chapter 6 summarizes the key uncertainties identified during these evaluations and recommends actions to resolve the uncertainties. XVll i i L.___.__ _ _ __
I. OVERVIEW 1.1 OVERVIEW OF THE MOLTEN SALT REACTOR EXPERIMEN T FUEL SALT DISPOSITION TASK 1.1.1 Facility Description The Molten Salt Reactor Experiment (MSRE) is an 8 MW reactor that was operated at Oak Ridge National Laboratory (ORNL) from 1965 through 1969 as a demonstration of the technology needed to develop a commercial Molten Salt Breeder Reactor (Robertson 1965). The reactor used a unique liquid fuel, formed by dissolving UF4 fuel in a carrier salt composed of a mixture of LiF, BeF , and ZrF
- The 2 4 fuel salt circulated through a reactor vessel, a fuel salt pump, and a primaiy heat exchanger at temperatures above 600°C (1112 °F). In the reactor, the salt was forced through channels of graphite to provide the geometiy and moderation necessal)' for a nuclear chain reaction. Heat was transferred from the fuel salt to the secondary coolant salt in the primal)' heat exchanger. The coolant salt is similar to the fuel salt, except that it contains only LiF (66%) and BeF2 (34%). The coolant salt passed from the primary heat exchanger to an air-cooled radiator, a coolant salt pump, and then returned to the primal)'
heat exchanger. Each of the salt loops was provided with drain tanks, located such that the salt could be drained out of either circuit by gravity. A single drain tank was provided for the nonradioactive coolant salt. Two drain tanks were provided for the fuel salt. The fuel salt drain tanks were provided with a system to remove the intense heat generated by radioactive decay immediately after an emergency reactor shutdown and fuel salt drain. A third drain tank connected to the fuel salt loop was provided for storing a batch of flush salt. This batch of salt, similar in composition to the coolant salt, was used to condition the fuel salt loop after it had been exposed to air and to flush the fuel salt loop of residual fuel salt and contaminants before accessing the reactor circuit for maintenance or experimental activities. As seen in Figs. 1.1 and 1.2, the reactor circuit is located in a large cylindrical cell. The coolant circuit, including the coolant salt drain tank, is located to the south of the reactor cell. The two fuel salt drain tanks and the flush salt drain tank are located in a rectangular cell north of the reactor cell. The coolant salt circuit, including the salt-to-air radiator, is located south of the reactor cell. Blowers force outside air through the radiator and up the large diameter coolant stack attached to the south exterior wall of the building. The design of the fuel salt drain tanks is shown in Fig. 1.3. The tanks were constructed ofHastelloy N (referred to as INOR-8 in the MSRE design documentation), as were all components that were. in contact with molten salt. An array of heat transfer thimbles penetrates the head of the fuel salt drain tanks. Bayonets, attached to a steam drum assembly, are inserted into these thimbles. Cooling water was fed through a tube inside the bayonets and evaporated by the heat transferred out of the drain tank; steam was collected in the steam dome. The flush salt tank is similar, except that it does not have a cooling system. Each of the salt drain tanks was suspended in a furnace assembly with resistance heaters provided to maintain the salt above the solidus temperature. Each tank is sized to hold the entire inventol)' of fuel or flush salt, respectively. Thus, there is one redundant fuel salt tank. The drain tank cell (Fig. 1.4) is at a low elevation in the facility so that salt could drain by gravity from the reactor circuit into the tanks. A system of freeze valves, in which cooling air was blown across *a flattened section of salt piping to form a plug of frozen salt, were used to isolate the reactor circuit from the drain tank system, the drain tanks from each other, and the drain tanks from the salt processing system. Salt was transferred back to the reactor circuit, or on to the salt processing facility, by freezing salt plugs in 1-1
1-2 0RNL-0WG 63*1209R REMOTE MAINTENANCE
/ 1 CONTROL ROOM REACTOR CONTROL ROOM I
J =- I. REACTOR VESSEL 7. RADIATOR
- 2. HEAT EXCHANGER 8. COOLANT DRAIN TANK
- 3. FUEL PUMP 9. FANS
- 4. FREEZE FLANGE 10. FUEL DRAIN TANKS
- 5. THERMAL SHIELD 11. FLUSH TANK
- 6. COOLANT PUMP 12. CONTAINMENT VESSEL Fig. 1.1. Arrangement of the principal components of the MSRE.
D D LI '1,,' MAINTENANCE STORAGE
~-**r;;i-**
EQUIPMENT 1* SPARE CELL
- NORTH
!ELECTRICAL SERVICE PRATICE CELL CELL AREA
,, INDUCTION REGULATOR , . CELL LIQUID WASTE CELL DECONTAMI* NATION CELL I l,J
.) AUXILIARY I CHARCOAL 'I BED
- l 0 1' 1 I
*I CHARCOAL BED CELL FUEL DRAIN I TANKNO.1 ! 36-ln. DUCT CONNECTING THE DRAIN TANK CELL j WITH THE REACTOR CELL BLOWER HOUSE --K--=
Fig. 1.2. Plan view of the MSRE facility. I I
1-4 UNCI.ASSIFIED ORNL-LR-DWG 61719 INSPECTION, SAMPLER, AND LEVEL PROBE ACCESS BAYONET SUPPORT PLATE STRIP WOUND FLEXIBLE HOSE WATER DOWNCOMER INSTRUMENT THIMBLE FUEL SALT SYSTEM FILL AND DRAIN LINE FUEL SALT DRAIN TANK TANK FILL LINE 0 THIMBLE POSITIONING RINGS Fig. 1.3. Sketch of an MSRE fuel drain tank.
IL j
~-- --- =ir HIGH BAY OFFICE OFFICE
~ . ~~
I TRANSMITTER
- I .... ~ EL 842"-0" ROOM EL 840"-0" I * ;4 : *
** ' ' - ~ *
. *. **' ...~***EL838" C.
~ .. ***.I**
* -~-.- -. ~ - ::
. -6. ..~,*.~ ~*-,.*:*_.**:
II ... I** : .. **~. --~.* ~. .... :*.i. <. :.-. NORTH ELECTRICAL
~-
\
* =-*' * ..t **
SERVICE AREA (NESA)
. EL 834'-6"
- . 0,
- DRAIN TANK CELL
.,
- 4 *
..4 14" *,
f"D-1
\
*I .. .... '* ..... . .. . .... .. .. . . . . . .. ....
~
.... * * -l
-~----** /'*/ ,* -;;* ~~ / :: ..
I : -~ ... : , * "".t ** ;. . Fig, 1.4. Cross-section through the MSRE drain tank cell and adjacent areas.
1-6 the appropriate freeze valves, closing valves in the off-gas system, and pressurizing the tank with helium. This forced salt through one of the dip legs reaching the bottom of the tank. The drain tanks are provided with a set of thermocouples on the side and bottom of the tank. In the case of the fuel salt drain tanks, an additional set of them10couples is located in a vertical thimble that extends down into the salt. The tanks are suspended in the furnace assemblies from weigh cells. The tank wall is 0.5 in. thick. The thimbles are constructed of l.5-in. Schedule 40 pipe. Each tank is provided with a 3 in. access pipe and flange at the top center. In the case of the fuel tanks, this pipe extends up through a gap in the center of the steam dome. The salt processing facility was constructed in the fuel processing cell, just south of the drain tank cell. This process (Fig. 1.5) was used to remove oxides and corrosion products from the salt by sparging with a mixture of H2 and HF and to remove UF4 by oxidation to UF6 with a F2 gas sparge. The sparging operation took place in the fuel storage tank, which was sized to hold the entire inventory of fuel salt or flush salt with enough freeboard that salt carryover would not be a problem during gas sparging operations. In the fluorination configuration (depicted in Fig. 1.5), the sparge gas flowed through a high-temperature NaF trap to remove volatile fission products; a bank of NaF traps in the high bay to remove the uranium; a caustic scrubber to neutralize the excess fluorine; and an array of mist, high-efficiency particulate air (HEPA), and charcoal filters. Most of the filter system is located in the spare cell. The reactor and drain tank cells were designed as a sealed containment system. Both are lined with stainless steel and are covered with two layers of shield blocks (Fig. 1.6). A stainless-steel seal pan between the t\vo shield block layers is welded to the cell liner, completing the containment boundary. The cells are designed for remote maintenance (Blun1berg and Hise 1968). A maintenance shield (Fig. I. 7) is used to access the cell. The shield consists of a frame and track assembly on which stationary and sliding concrete shield elements are placed. Some of the shield elements have rotating work plugs, with tool bushings and viewing ports. To place the shield, the appropriate upper shield blocks are removed and the seal pan is cut away to expose the lower shield block. The track and shield blocks are set up around the lower block, and the operators move to the shielded remote maintenance control room (or a similar safe area). The lower block is lifted out through the opening in the frame with the building crane. A motor is used to move the sliding shield element over the gap~ this allows the operators to reenter the high bay. The shield elements are then positioned so that tools can reach the component being worked on. The shield was also used to access the drain tanks by placing a module through one of the shield ports that couples onto the flange on top of the drain tank. The arrangement that added most of the 233 U to the fuel carrier salt is shown in Fig. 1.8. 1.1.2 Operating History Production of fluoride mixtures for the flush and coolant salts began in March 1964 and was completed in September (Shaffer 1971 ). These salts were used in the checkout of reactor and coolant systems, and the flush salt was purified by using the MSRE fuel processing system after initial circulation in the reactor circuit (Lindauer 1969). Processing of the fuel carrier salt began in November 1964 and was completed in March 1965. The original uranium charge, consisting of30% 235 U and 70% 238LJ, was then added to the fuel carrier salt. Although the design documents indicated that salt would be added from a station in the high bay and through components of the fuel processing system, it was actually added directly into the drain tanks from an addition station set up on top of the drain cell shield blocks. The tanks were hot when sail was melted in furnaces and fed into the tanks. During the operations phase of the MSRE, the salts were maintained in the molten state.
I ORNL-DWG 68-8994 Ii SALT rrf.200°F NaF ABSORBERS IN CUBICLE ~-I SAMPLER ~ l{r===i I CH~~~TNG HIWE8ty ij == == ~ 7r ===u-I ___ , O ,~ SALT TO OR He Af Dl
~ .----,
FUEL PROCESSING CELL ..........a n U TO
._.LIQUID ACTIVATED CHARCOAL TRAP WASTE1A FROM DRAIN AND FLUSH
~ MIST TANK rs TO VENT SYSTEM
- ,j TANKS STORAGE ~ FILTER . * ,_.
I
--.,J i n SODA I
U LIME TRAP WASTE 750°F CAUSTIC CHARCOAL TRAP SALT NaF BED NEUTRALIZER TANK II Fig. 1.5. Flowsheet of the MSRE salt processing system. ('
1-8 Unc:lassi:f'ie:i
. ,,.,.. 'G oru- 64 -::o;.
.{.(
/;,--
1 I I I I -3~ ...... A I. 2 --4 I I~~
,&!'I I I I I: ~.8 1*I I I I
I I la:> I 4
"7
!I I
5 8
- I iI
- I 6
9
~:I I
!l I
C
~h IN, I I I I I I I ;; I ,
X" I I D "xjX"
.J I
I I 10 11 ~ I 12 I ,I ., t
.J
~I= I
=I:
- _ II I I I I : I I ! I E =-=:-,;
lo,..( l'*= I I 13 I 14 15 I
.: ~ = I I I (\ii ::: ~~
~Iii !I I I iii If) I F .J I I I 16 :I 17 I I 18 I I I I II ,,I .II I
; I I '
lol tlf> 1I 1 I 19 i I 20 II I I 21 '" I l'I I If) 1 lf)I 1 H G
~
I 22 -: ;I iI 23 I I I I 24 --t 1}' L,/, I
"~ 25 I 26 I I 21 __ lv J
PLAN 1~:
.ii~
W.~LL~ Fig. 1.6. Shield block arrangeme nt on top of the drain tank cell.
1-9 ORNL-OWG 67-13761
*--{t--MOTOR DRIVE (DRIVE NOT SHOWN)
FRAME OVER DUAL
\ LOWER ROOF-PLUG OPENING ACCESS OPENING FOR LARGE TOOLS AND /
7'
\
' ~WORK PLUG MAINTENANCE CONTROL ROOM
.f'---------./-------~l-------~v' i_ Fig. 1.7. Typical plan view of setup of the maintenance shield.
1-10 ORNL-DWG 68-967
'lll---T URF CARRIER GRAPHITE SAMPLING SHIELD
'i-
*t* **. i:
**r~r=MBLY CONTAINMENT ENCLOSURE r 1 Fig. 1.8. Arrang ement of the equipment used to add 233 0 directly into an MSRE fuel drain tank.
1-11 The reactor achieved initial criticality on June 1, 1965, and reached full power early in 1966. In 1968, the initial charge of uranium was stripped out of the fuel and flush salts by using the fluoride volatility process (Lindauer 1969). Residual uranium was 7 ppm in the flush salt and 26 ppm in the fuel salt. Uranium recovered from the salt was sent back to the uranium enrichment facility in Portsmouth, Ohio. Each batch of salt was then sparged with an HF/H2 mixture to precipitate corrosion products that resulted from the fluorination process, and the salt was filtered as it was returned to the drain tanks. A second fuel charge, consisting of -80% 233 U and 20% 235U, was then used for the balance of reactor operations. The uranium concentrate salt was again added directly into one of the drain tanks by using the equipment depicted in Fig. 1.8. Late in the operating campaign, the fuel was enriched with relatively small quantities of 239Pu (as PuF3). In December 1969, reactor operation was terminated, and the fuel, flush, and coolant salts were drained into their respective drain tanks. The fuel salt was divided between the two fuel salt drain tanks. As the fuel salt was drained, activity was observed in the drain tank cell atmosphere. This activity was traced to a minor leak in freeze valve l 05, leading to fuel drain tank 2. All of the salts were allowed to freeze, but the fuel salt was held at an elevated temperature for about a year to control the effects of radiolysis in the salt. This is the only time the salts were allowed to freeze in the MSRE drain tanks, and they have not been melted since. Throughout the operation of MSRE, a detailed record of the chemical behavior of the salts was maintained, and a summary of the chemical aspects of MSRE operation was prepared following the postoperation examination period (Thoma 1971 ). 1.1.3 Postoperation History In the 2 years that followed reactor shutdown, a number of activities were performed to evaluate the performance of materials used to construct the reactor and to improve the containment of the salts stored in the drain tanks (Rosenthal ct al. 1971). The latter included the isolation of salt and off-gas lines between the drain tank cell and tl1e reactor system and removal of the failed freeze valve (and adjacent freeze valve 106) for examination. Mechanical devices were used to press copper plugs into the end of the severed salt lines. These plugs were also provided with leak detection lines. During a series of experiments to test the chemical compatibility of metal and graphite materials with the MSRE salts, the production of fluorine gas was observed (Briggs et al. 1964). It was found that irradiation of solid salt produces fluorine radicals, which can then combine to form F2 and diffuse to the salt surface. Some of the irradiation experiments were reconfigured to explore this phenomena. It was found that lower temperatures favored the fom1ation of F2; whereas at higher solid salt temperatures, the back reaction of fluorine with the lithium and beryllium from which it had been dissociated was favored. No uranium was observed in the gas above the salt. To prevent the accumulation of fluorine in the drain tanks after the reactor was shut down and the salts were solidified, a procedure was instituted for an annual annealing of the fuel salt to prevent tl1e release of radiolytic fluorine from the fuel salt (Guymon 1971). In the annealing operation, the salt was heated with the furnace heaters to a temperature above 149°C (300°F), but well below the melting temperature, and held at tl1is temperature for about a week. Following reactor shutdown, the reactor operations crew was disbanded, and monitoring of key MSRE parameters was established at the central ORNL waste management center. A number of studies addressed the disposition of the fuel and flush salts and the decommissioning of the MSRE facility. Final disposition of the fuel and flush salts was put on hold pending the opening of a geologic repository. In 1985, a review was perforn1ed of the issues associated with extended storage of the fuel and flush salts in their drain tanks (Notz 1985). This review extended the radionuclide decay models out to repository time frames and included a review of fluorine production predictions. Additional salt irradiation experiments were carried out to provide a better basis for fluorine generation predictions (Toth and Felker 1990).
1-12 Over the years, indications were noted of migration of radioactive materials in the piping connected to the drain tank cell and the off-gas system. In particular, increased radiation levels were observed in the north electrical service area adjacent to the drain tank cell. In addition to electrical penetrations, this service area houses components of the helium addition system that had been used to pressurize the drain tanks for salt transfers. Maintenance activities on the off-gas system indicated the presence of alpha activity. Gamma scans indicated the presence of 208TI, a product of the decay ol3 2 U. The annual annealing procedure was halted ruler December 1989 in part because of concerns that this procedure may have aided the migration of radioactive material out of the drain tank cell. In March 1994, a gas sample was taken from the auxiliary off-gas system, which was connected to the gas space above the drain tanks. Expectations were that fluorine might be detected in part per million quantities. An analysis of the gas (Table 1.1) revealed that fluorine composed about half of the gas sample (Williams et al. 1996). Even more unexpectedly, UF was a major component of the gas 6 stream. At a partial pressure of 69 mm Hg, the amount of UF6 identified was just below the 79 mm Hg saturation pressure ofUF6 at the sampling temperature of21 °C (69.8°F). Table I.I. Results from analysis of Molten Salt Reactor Experiment off-gas samples taken in 1994 Component Partial pressure (mm Hg) F2 350 Ine11s 305 UF6 69 MoF6 JO CF4 5 HF 0.74 N-F compounds trace The auxiliary off-gas system was in communication with the vapor space above the drain tanks; this provided a migration path for UF6 and F2 that had been liberated from the salt. A detailed review of the off-gas system (Fig. 1.9) was undertaken to identify all locations to which uranium might have migrated. It was also recognized that the amuliruy charcoal bed was in communication with the drain tanks because valve 561 had failed in the open position. Radiation and thermal surveys were used to identify a uranium deposit in the auxiliary charcoal bed (Fig. 1.10). These data indicated that 2.6 kg of uranium was deposited in a I ft section of bed near the inlet pipe. It was also discovered that reactions of fluorine with carbon at low temperatures can result in the formation of potentially reactive carbon-fluorine compounds. Laborato1y investigations determined the likely stoichiometry of these compounds to be that of C F. On 2 the basis of volume estimates of the piping and vessels in communication with the drain tanks and the measured concentration ofUF6 in the off-gas, at least 1.8 kg of uranium is believed to exist as vapor in the vessels and piping. Because the measured UF6 concentration is so close to the saturation pressure, solid deposits of UF6 may exist in the off-gas system. Thus, the amount of uranium present in the system could exceed 1.8 kg.
r-~:! ~ox -----J:ltV -H-
-r---- AREA38 - - o~~)
I I SE 1 AREA ,_ - HSY. I )( 1171A XIHSV,. fflA I I
- - .I I fC I )( ~~j l--- f" I
I I HCV-1 1178E X I rN I I I HalUM SUPPLY I MV I (REMOVED) I I I I 1AREA 1 L-'---.L --, 1 2S 1 J. ______ ______ _ _ V-5
--------------LL.I V-562C1 NORTH ELECmtc SERVICE AREA V-562C3 V-562C2
~ . .Ttm TO FILTER v~ AND STACK V-522A r-- V-557B
'**1.I HCV-577 REACTOR --I
\ .,)
V-522B V-524C CV-524 .1. I FV-104 FV-103 E-M I I TOFST I I I I I I I REAC10R CELL I r--------------------------J L----------------------------------------- VENT HOUSE : CHARCOAL BED CELL : X FAIL CLOSED FUEL OFA/LOPEN FLUSH TANK DRAIN TANK CELL Fig. 1.9. Princip al components of the MSRE off-gas system.
iJnai1tflh~! DEPTH TOP EL.840'-G "
~VEHrsn5ust SCALE SHIELD BLOCK (In,) 111 In, TlllCK EL 11.c11*-o*
TEMPERATURE ( 0 r) 12
}
61 62 63 n-------- 60 64 65 66
,=::------- -
67 68 69 70 24 20 36 ~1'f-9"1 OUND ~
- FAS *~.
~ . . . (LOCATION or
; VALVE VS51) 48 60 7/20/94
- i'ffi\m =1 ., .* 25 30 \ ~
72 84 11/4/94 +;'
" \
35 40 ' ............ 961 108 120 24'-3"
-t. ,
In,-
} i~ ~f B 1
*--------\----45_
I 50
' .\
-----------~---------
)'
132 ...J >- ,. ..
- j DCPOSIT r~~Y-*~1-'. , 55 '/
- f:t~;~~~i W c, 144 ~g -~**, i~-~- ------~----- /.
e ;;*
...,o ~
,**I
_ -,r- ---------- 156 o::Z wo 1-0::
*.* . In.- ,.- :c 65 ._
168 <:r
- i:u ; ~::.- ... I 180 ., ~2-a.
w 70
. .i:,.
192 ,.*... ,*
. C
/
75 204. 216 1 LI:.:*. I 80 I 2281 I*, ~.. : 2.40 f
- l. *.: .. -, 85 -
* - RADIATION SIDEWALL TEMPERATURE t..** AUXILfARY ., 90 INrRARED ,. l I
I.*, ' 2s2 CHAl?CCAL RED r* SIDEWALL TEMPERATURE 95 SPLIT JCT 276 EL 82f-9'
...~ j 100 28R 1/2 - 0 2 A. 6 8 10 12 14 16 18 20
~ .: *r*~.\**:.~ ~*- :<-~--~. ~-~*: ., DOSE RATE @I fl IN WATER (rem/h) i I
I Fig. 1.10. Location of the uranium deposit in the auxiliary charcoal bed.
1-15 1.1.4 MSRE Remediation Project 1.1.4.1 Organization of the MSRE Remediation Project Following this discovery, the MSRE Remediation Project was initiated to assess, control, and reverse the consequences of the migration of uranium. These activities are being planned and executed according to the requirements of the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA). In November I 989, the U.S. Department of Energy (DOE) Oak Ridge Reservation was placed on the National Priorities List, and a Federal Facilities Agreement (FFA) was established between DOE, the U.S. Environmental Protection Agency (EPA), and the Tennesse e Department of Environment and Consen1ation (TDEC). A 1995 revision to the FFA established that the MSRE Remediation Project is to be conducted and documented in accordance with the requirements of CERCLA. This ensures that a risk-based process is used to identify appropriate actions and that both the state of Tennessee and EPA Region IV are involved in the decision making process. The MSRE Remediation Project is organized into three CERCLA projects. The first of these, a time-critical removal action, is further divided into two sets of activities (Lingle 1995). The first, called interim correctiv e measures, addressed various aspects of containment, criticality control, and protection against chemical reactions. This phase of the project, discussed further in Sect. 1.1.4.2, is now complete. The second phase is trapping uranium and fluorine in the off-gas system, converting the uranium to oxide, and placing it in storage. These activities are discussed in Sect. 1.1.4.3. The second CERCLA project, classified as a non-time critical removal action," will remove the deposit of uranium that has been trapped on the m1\'.iliaiy charcoal bed. It is summarized in Sect. 1.1.4.4. The third CERCLA project is the fuel salt disposition task, which is the subject of this report. The fuel salt disposition task is classified as an interim remedial action. This classification defines the CERCLA documentation that must be prepared before conducting remedial activities. In particular, a feasibility study is being prepared to define the risk and the remediation goals, to identify the remediati on alternatives, to screen potential remediation technologies, and to conduct an overall evaluation of the alternatives. The feasibility study will be followed by a proposed plan, which details the specific activities to be used to conduct the task. The selection process will culminate in the issuing of a Record of Decision, agreed to by all of the parties in the Oak Ridge FFA. Uraniwn left in the MSRE following reactor shutdown also has been included within the scope of the Defense Nuclear Facilities Safety Board (DNFSB) Recommendation 94-1, which addresses material left "in-pipeline" following shutdown of nuclear facilities without complete decommissioning activities The implementation plan for this recommendation carries the milestone that the salts will be removed from the drain tanks by May 2002. The implementation plan also states that the uranium will be separated from the salt and converted into an oxide form for long-term storage. Completion of these activities is contingent on the remedial action alternative documented in the CERCLA Record of Decision. ""Action Memorandum for Uranium Deposit Removal at the Molten Salt Reactor Experiment, Oak Ridge National Laboratory, Oak Ridge, Tennessee," June 1996.
1-16 1.1.4.2 Interim corrective measures At the time the uranium deposit was discovered in the auxiliary charcoal bed, the concrete charcoal bed cell in which the main and auxiliary charcoal beds are located was flooded with water. This water was removed to keep the potential neutron moderator out of the charcoal beds. Water sources throughout the facility were either removed or isolated from potential uranium deposits to prevent a nuclear criticality. Reaction of the carbon-fluorine compounds in the charcoal bed cell could lead to a significant release of energy that could rupture the auxiliary charcoal bed vessel and generate a pressure in the cell to lift the concrete cell lid and disperse radioactive materials into the environment. To prevent this, a lid restraint ring was installed. The cell was inerted \\~th CO2 so that, should the carbon-fluorine compounds react, the burning of charcoal in air will not add to the internal pressure in the cell. Valve 561, which had previously foiled in the open condition, was repaired and is now closed. This isolates the auxiliary charcoal bed from the drain tanks. New, calibrated pressure transmitters were installed on the heliun1 lines leading from the north electric service area to the two fuel drain tanks (PT 572 and PT 574 shown in Fig. 1.9) and on the off-gas line downstrean1 of the auxiliary charcoal bed (PT 562). Along with a transmitter installed previously upstream of the charcoal bed (PT 518) and another installed as part of the reactive gas removal system (PT 542), these instruments allow the trending of pressures throughout the system. Readings are as high as IO psig and indicate that plugs exist in the system between the drain tanks and the reactive gas removal system and between the reactive gas removal system and the branch of the off-gas system ahead of the charcoal bed cell. Before closing valve 561, it was observed that readings upstream and downstream of the charcoal bed cell (PT 518 and PT 562) were identical. A new, computerized data collection system was installed to record these pressures, along \\~th temperatures and other key data points throughout the system (including drain tank and drain tank cell temperatures). A general review of facility data collected over the past few years was performed and documented (Shor 1996). A detailed survey of the rest of the facility was performed to identify any other potential migration paths for uranium or other radionuclides. Gamma radiation from 208 TI, a decay product of 232 U, was identified in the fuel processing cell and a salt distillation experiment located in the spare cell. Apparently, the freeze valves on the salt lines do not provide a gas-tight seal, allowing migration of the pressurized gas above the tanks into these areas. An analysis of data collected in these cells indicated that the total amount of uraniun1 present is <700 g and possibly as little as 20 g (DeVore et al. 1996). 1.1.4.3 Reactive gas removal The sampler-enricher hardware, used during reactor operations to withdraw salt samples for analysis or to add enriching snit to a well-mixed location in the pump bowl, also provides an enclosed access point to line 542 in the off-gas system. A glove box facility has been attached to the sampler-enricher enclosure to allow the wiilidrawal of gas into a trapping system. The major components of this system are shown in Fig. 1.11. A sodium fluoride trap is used to remove UF from the MSRE gas, 6 followed by an alumina trap for the removal of fluorine. A molecular sieve trap is used to trap residual moisture, possibly containing HF from the alumina trap, and to prevent corrosion of other system components. Gas from tl1e traps is sent to holdup tanks for the decay of 220Rn and its daughters before it is discharged into the MSRE ventilation system through fluorine-resistant HEPA filters. A Fourier-transform infrared spectrometer system is used to monitor UF6 concentrations between the NaF and alumina traps and at the outlet of the trapping system. Thermocouple data and gamma transmission measurements are used to monitor the loading of the traps. As of the writing of this document, the
FilR-1 FflR-21 Computer Process Mon*tbr*dg Computer Cornput~j IR-1 Cell (10 cm) 1 Process Ccmputational Multiplex~~ _r1 IR-3 Cell {4 111)
~,
1 Process Gas y I ToMPX Instruments
&Controls Process Gas-+
Computer ToMPX To Monitoring Instruments & Transducers i' ToGanuua Detectors
~
R-2
~::~--11~ ~-c,.
f -~~--- .:! t Pr~~ss
**1
- I, Input from MSRE Sampler/ t Process Oas-+
t Process Gas -+
. -* *--~
Enricher Box. I I=--__ Thermocouple assembly Holdup ( S (S thermocouples) Tank or Vacuum Piuup u1 e
;:. ;s enu *-
a::E "B ,-~---r~ - HEPA "3 ~
-~ c.. I*- I Filter System I
--.J
~it
~
~Co--
- _e
+
To HVAC Molecular Sieve Duct Gamma Source Detector UF6 Trap Fig. 1.11. Flowsheet of the reactive gas removal system.
1-18 reactive gas removal system is nearing operation. When the uranium has been trapped on the NaF, it will be placed in temporary storage until a conversion facility is constmcted. This facility will be used to convert the UF6 absorbed on the NaF to U3O8 for long tem1 storage at ORNL. A report recommending the uranium conversion flowsheet has been written (Advanced Integrated Management Services 1995), but design of the conversion facility has not yet begun. The reactive gas removal system will be operated in batch mode by evacuating the holdup tanks downstream of the traps to a pressure that will allow a predetennined amount of gas to flow through the traps. As batches of process gas are removed, the off-gas system will drop below atmospheric pressure. Eventually, the off-gas system will be evacuated and backfilled with helium. Initial operation of the reactive gas removal system ,,1ll remove process gas from that part of the off-gas system that is not in some way isolated from line 542. As noted earlier, pressure differentials as high as 8 psi occur between segments of the off-gas system, indicating that the piping is plugged in some way. Because the UF 6 partial pressure is so near saturation, the system may be plugged with solid UF6* If so, evacuating the system will cause the plugs to sublime, and access may be gained to the rest of the system. If not,jumper lines may be installed to access other parts of the system. Line extensions may also be installed to remove gas from the salt processing system and distillation experiment in the fuel processing and spare cells. 1.1.4.4 Charcoal bed uranium deposit removal Studies are now underway to establish the performance of an approach to passivate carbon-fluorine compounds by flowing a reactive gas through the auxiliary charcoal bed. If the studies yield a gaseous treatment approach, little additional equipment will be required for the passivation operation, and risks during other operations will be reduced. After the charcoal is removed, the container will be transferred to a hot cell where the contents will be divided into several containers suitable for temporary storage. The charcoal will be placed in interim storage until the uranium conversion facility is available and will then be converted from UF4 to U O 3 8 for long-tenn storage. Conversion is expected to take place in the same facility used to convert the UF 6 collected by the reactive gas removal system, with appropriate modifications made to the head-end of the process. 1.1.5 Fuel and Flush Salt Disposition Task The fuel and flush salt disposition task addresses the activities needed to place these salts, and the key contaminants in the salts, in an appropriate condition that provides for the long-tenn protection of human health and the enviro1m1ent. The primary objectives of the fuel salt disposition task are to reduce risk as identified in the CERCLA process and to stabilize material identified as part of DNFSB Recommendation 94-1. Secondary objectives are to reduce or eliminate maintenance and surveillance costs associated with storing the snits in MSRE and to allow the final decommissioning of the MSRE facility to proceed as soon as funding is available for the task. The key contaminants include fissile materials, radioactive materials, and any chemicals in the snits or salt by-products that are identified as posing a significant risk to human health and the environment. The scope of this task clearly includes the salt in the fuel and flush drain tanks. The evaluation currently underway will establish the extent to which residual salt in those MSRE systems that have contained salt are in the scope of this task, including residual salt left in the drain tanks after removal. This task does not include the ultimate decommissioning of the drain tank cell or ofMSRE as a whole. Locations of residual snit \\1ll be defined:* and recommendations will be made on whether residual snit
1-19 in the reactor, drain tank, or salt processing and experiment systems will be removed and disposed of in this phase or as part of the ultimate facility decommissioning. This task will assess whether a truly pemmnent disposition cnn be achieved in the near term. Ideally, the snits nod the key contaminants in the snits should be placed in an appropriate permanent disposal site or should be made available for productive reuse. If it is not possible to commit to permanent disposal at this time, interim storage must be shown to provide adequate protection to human health and the environment. Definition of credible end points is discussed in Chap. 2. At the start of salt disposition operations, the reactive gas removal system will have purged out the off-gas system, nod the amount of urnniwn nnd fluorine removed from the off-gas system will be known. In the course of evacuating the off-gas system, which includes the upper part of the drain tanks themselves, the integrity of the system will become better defined. Significant leaks will be identified if the system cannot be fully evacuated. Plugs in the off-gas system will be identified and likely removed, and the operating condition of other valves in the system (such as the off-gas system valves in the drain tanks) may be established. The last pressure transducer in the north electric service area, attached to the helium line leading to the flush salt drain tank, will be replaced. The charcoal bed deposit removal operation will also be well undenvay, if not completed. The amount of uranium and fluorine in the charcoal bed should be quantified. Even if operations at the charcoal bed cell are not complete, they are physically isolated from activities at the drain tank and salt processing cells. This report is pnrt of the preparation of a CERCLA Feasibility Study. The CERCLA document is scheduled to be transmitted to TDEC and EPA Region IV at the end of February 1997. A conceptual design of the recommended alternative will be prepared next and will serve both as the basis for the CERCLA Proposed Pinn and the planning base budget for the project. Studies into the behavior of irradiated salt have also begun and will continue into next year. Work to access the drain tank cell, and possibly inspect the interior of a fuel salt drain tank, will also begin next year. Assuming the salt is to be removed, current plans call for removal operations to be completed by 2002, and separation, stabilization, and storage activities to be completed in the following year. 1.2 CHARACTERIZATION OF THE MSRE FUEL AND FLUSH SALTS 1.2.1 Inventory and Compositio n of the Fuel and Flush Salt The coolant and flush salts were initially prepared as 66 mol % 7LiF and 34 mol % BeF , using 2 isotopically pure 'Li. In the case of the coolant salt, no changes to the composition occurred during reactor operation. Although some contamination with tritium is likely as a result of diffusion from the fuel salt across the hot heat exchanger tubes, this salt remains as a stable, largely nonradioactive solid in the coolant salt drain tank. The carrier salt for the fuel was initial prepared as 65 mol % 'LiF, 30 mol % BeF2, and 5 mol % ZrF4* Zirconiun1 fluoride was added to the fuel salt so that if the salt became contaminated with O>..")'gen-containing contaminants (e.g., water or oxygen), zirconium oxides would precipitate ahead of uranium oxides. Enough zirconium was added so that the precipitate would be observed before precipitation of uranium began. Fuel was added to the carrier salt as a 73 mol % 'LiF - 27 mol % UF4 low-melting eutectic enriching salt. The bulk of the fuel ,,,as added directly into the molten carrier salt in one of the drain tanks. Final approach to criticality was made by adding small batches of enriching salt to the fuel nod carrier salt using the sampler-enricher, while the salt was circulating through the reactor. The initial 235 U fuel charge was diluted with about 70% 238 U lo increase the amount of uranium in the salt until
1-20 experience was gained with maintaining chemical control of the system, especially with respect to maintaining the desired UF/UF4 ratio. The total mass of uraniwn in the initial charge was about 220 kg. The composition of the fuel salt did not remain constant. The quantity of uraniwn changed as fuel was burned and additional enriching salt was added. Enriching salt also added small amounts ofLiF to the system. Plutoniwn was produced by transmutation of 238 U. Each time the system was opened for maintenance, flush salt was circulated through the reactor circuit before the fuel salt was returned. It was estimated that about 20 kg of salt remained in the reactor circuit following a drain; this amount of fuel salt was lost to the flush salt and the same amount of flush salt was added to the fuel salt each time the system was opened. In 1968, the initial charge ofuraniwn was removed by fluorination, down to 26 ppm uraniwn. A second charge of uraniwn, about 80% mu and 20% 235 U, was then added. By this time it had been detennined that uranium chemistry control was not a major concern, and no 238 U diluent was used. The total charge of uranium used for 233 U operations was about 40 kg. Again, bum up, additions of enriching salt, and cross-contamination with flush salt contributed to minor changes in salt composition. Late in the 233U campaign, enriching salt containing :?39JluF3 was used to demonstrate reactor operation using that isotope. Changes in the composition of the flush salt are similar to those in the fuel salt, except that burn up and the addition of enriching salt do not apply. Seven cross-transfers with fuel salt occurred during 235 U operations. The flush salt was fluorinated in 1968, and then returned to its drain tank with a heel of 7 ppm uraniwn. Two more cross-transfers took place under mu operations. Thus, the fraction of uranium that is 235U in is greater in the flush salt than it is in the fuel salt. Throughout reactor operations, a meticulous sampling and analysis program tracked both the fuel and flush salts. Data from this program is summarized in Thoma (1971) and has been reviewed by Williams et al. ( 1996). The latter includes estimates of the effects of cross-contamination of the flush salt each time the reactor was opened for maintenance. The final mechanism that changed the composition of the fuel salt is the production ofF and UF 2 6 by radiolysis during the time the salt was stored as a radioactive solid. Because the amount of UF 6 present in the off-gas system (especially the amount that may be present as solid deposits) and in the auxiliary charcoal bed is only approximately known, this represents the major uncertainty in the present composition of the fuel salt. It is assumed that loss of uranium from stored salt is an issue for the fuel salt only. However, it is also possible that the flush salt could contain deposited UF6 on the salt surface that migrated over from the foci salt. The current estimates of the inventory and composition of the fuel and flush salts stored in the drain tanks are given in Tables 1.2 and 1.3. 1.2.2 Fissile Material Inventory As noted in the previous section, fissile material inventories were meticulously tracked by a program of sampling and analysis throughout reactor operations. Thus, inventories in the fuel and flush salt discharged into the drain tanks in December 1969 were very well docwnented. The major uncertainty is the amount ofUF6 that has left the fuel salt as a result of radiolysis. The fissile material inventories in the fuel and flush tanks, both with and without the migration of 4.4 kg of uranium from the fuel salt, are shown in Table 1.4.
1-21 Table 1.2. Inventory of salt stored in the fuel and flush drain tanks Tank Mass Volume Density (kg)" (ml)° (g/mL at 26°C) Fuel salt Fuel salt drain tank I 2479 1.00 2.48 Fuel salt drain tank 2 2171 0.88 2.48 Total fuel salt in drain tanks 4650 1.88 Flush salt Flush salt drain tank 4265 1.92 2.22 "Mass and volume estimates that best correspond to process history. Table 1.3. Composition of the fuel and flush salts Fuel salt Flush salt Current salt composition, mo/% 64.5 65.9 30.4 33.9 4.9 0.18 Mass offissile elements in salt, kg Uranium <33.2" 0.5 Plutonium .724 .013 Current distrihulion <">/fissile element isotopes, w/ % 232 0 160 ppm 75 ppm 233 0 83.92 39.4 23 u 7.48 3.6 2350 2.56 17.4 236 0 0.104 0.245 23 8U 5.94 39.4 239 Pu 90.J 94.7 24
°Pu 9.52 4.8 other Pu 0.35 0.50 0
Assumes that at least 1.8 kg of uranium has migrated to the off-gas system, and at least 2.6 kg of uranium is loaded onto the auxiliary charcoal bed. A swnmruy of key criticality safety parameters, including the always safe" mass, is given in Table 1.5. The fuel salt, of course, contains sufficient 233 U to achieve criticality. The total mass of235 U and 239 Pu also exceeds the always safe mass. In evaluating criticality safety, the combined effects of all fissionable isotopes must be considered. If all uranium is removed from the fuel salt, the 0.65 kg of 239Pu is still above the always safe mass. However, any further division of the fuel salt (such as into the two drain tanks) results in a 239Pu invento1y below the always safe mass. In the case of the flush salt, the sum of 233 U, 235 U, and 239 Pu is about 0.3 kg. This is well below the always safe mass for any of the three isotopes, and criticality is not an issue with the flush salt unless a significant quantity of uranium is found to have migrated over from the fuel salt.
1-22 Table 1.4. Mnss of uranium anti plutonium in fuel anti flush salt, kg Assuming no losses Assuming 4.4 kg U lost from fuel salt' Isotope Mass in Mass in Total Mass in Mass in Mass in fuel salt flush salt mass fuel salt tlrain tank tlrain tank I 2 Total U 37.I 0.5 37.6 32.7 17.4 15.3 b 232U mu 31.1 0.2 31.3 27.4 14.6 12.8 234U 2.8 0.02 2.8 2.4 1.3 I.I lJSU 1.0 0.09 1.0 0.84 0.45 0.39 236U 0.04 0.001 0.04 0.03 0.02 0.02 238l.J 2.2 0.2 2.4 1.9 1.0 0.9 Total Pu 0.72 0.013 0.74 0.72 0.39 0.34 ll9PU 0.65 0.013 0.66 0.65 (l.35 0.30 24°Pu 0.069 0.0006 0.070 0.069 0.037 0.032 Other Pu 0.003 0.0001 0.003 0.003 0.001 0.001 "Based on transport of2.6 kg to the auxiliary charcoal bed and presence of 1.8 kg as UF6 vapor in the off-gas piping and vessels. 6 As of 1995, there was 160 ppm total uranium in the foe! salt and 75 ppm total uranium in the flush salt. The 232 u,mu ratio is assumed to be the same in both the fuel and !lush salts. Table 1.5. Some key criticality safety parameters for fissionabfo isotopes present in the MSRE salt' "Always safe" mass 500 g 700 g 450 g Solid metal sphere, reflected with water 6 kg Aqueous solution ofUO 2F2 or Pu(NO3) 4 540 g 760 g 480 g Safe diameter for solution of UO 2F2 or Pu(N0 3) 4 10.5cm 13.7cm 15.4cm 0 So11rce: ANSI/ANS-8.1 Later in this docwnent, a package size that divides the fuel and flush salt into 24 packages will be considered. A package containing fuel salt (assuming 4.4 kg lost from the total fuel salt inventory) would contain about 1.2 kg of 233 U, 36 g of235 U, and 27 g o.r39 Pu. Only the 33 U would raise a criticality concern, even if these packages are grouped in sets of three. Even without uranium separation, the 8 g of 233U in a flush salt package would be of no concern. 1.2.3 Radionuclide Inventory Fission product inventories have been estimated using computer codes that track bum up and decay, especially the ORIGEN code. Decay products from uranium and transuranic isotopes are tracked using the fissile material analyses above and computer codes to track decay. Transfers of radionuclides between the fuel and flush salts are tracked in the same way as transfers of fissile material, and fissile material analysis provides supporting infonnation for estimates of material transfers between the salts. Because the MSRE was a fluid-fueled reactor, volatile fission products were not retained in the fuel. Initial estimates of the fission product inventory considered the transport of noble gases, iodine, and other volatile elements to the off-gas system. Since in many cases these transport mechanisms can only
1-23 be estimated, an additional w1certainty is introduced to inventol)' estimates for radionuclides with volatile precursors. In some cases, measured data was available to compare against predictions (Williams et al. 1996). Table 1.6 lists the half-life and inventol)' of those fission and activation products that are present in quantities greater than 0.1 Ci, as of 30 years after reactor shutdO\vn (about the beginning of the year 2000). It is seen that the inventol)' is dominated by 90Sr, 137 Cs, and their daughter isotopes. There will be a total of 24,400 Ci of fission and activation product isotopes at the start of the year 2000. Table 1.6. Inventory of fission and activation product isotopes" Isotope Half life Curies Inventory in fuel salt Inventory in flush (98.3%) salt (1.7%) 90 Sr 28.5 y 6,670 6,557 90y 113 2.7 d 6,670 6,557 113 137 Cs 30y 5,600 5,505 131mBa 95 2.6111 5,290 5,200 90 151 Sm 90y 117 I 15 2 147pm 2.62y 13.4 155 13.2 0.2 Eu 4.96y 4.4 4.3 154 0.1 Eu 8.8y 3.1 3.0 is2Eu 0.1 13.3 y I.I I.I 0.0 99 Tc 2.1 X J0 5 Y 0.5 0.5 0.0 12ssb 2.73 y 0.3 0.3 0.0 93Zr J.5 X J0 6 y 0.3 0.3 0.0 Total 24,400 23,985 415
'Bnsed on decay to December 1999.
Actinide radionuclide inventories are dominated by the 232U and mu decay chains and a few other key isotopes. The 232U and 233U decay chains are shown in Fig. 1.12. The 232U chain is a single chain until near the end, where it takes two paths to the end point of 208 Pb. One of these branches includes 208 Tl, which emits a distinctive 2.6 MeV gamma. The 232U chain also includes 220Rn, a gaseous isotope with a 56 s half-life. The 233U chain is a bit more complex, but the two branches have yields of only l % and 2%. Thus, only the main chain contributes isotopes with activities near I Ci at the present time. Radon is only present in the l % branch, and thallium in the 2% branch. The parent uranium isotope for the 232 U chain has a 70 year half-life, and the thorium daughter has only a 1.9 year half-life. Thus, this chain of daughter isotopes is already in equilibrium with the parent, and is decaying with the half-life of the parent. The parent uranium isotope for the 233 U chain has a 1.6 x lCf year half-life, and its thorium daughter has a half-life of 7,300 years. Thus, the thorium daughter has yet to build to equilibrium, and decay of the parent is not noticeable. Daughters of the 229Th daughter are short-lived, and are in equilibrium with 229Th. This daughter chain is only radiologically significant at present because of the relatively large quantity of the 233 U parent in the salt. Each of the other uranium and transuranic isotopes decays to another long-lived isotope, and their daughter chains are not radiologically significa nt compared to 232U and mu. The inventory of the significant actinide isotopes is shown in Table 1.7. Table 1.8 lists those radionuclides present in significant quantities in the salt along with a general factor for dose at l m from an unshielded point source of l Ci. It also lists the activity of each nuclide in a postulated package, in which the fuel and flush salts are each divided into 24 packages. The dose
I-24 U233 OR . . . . U233 99.99%-a 1.e-003%-SF U232 OR . . . U232 99.%-a 1.%-SF Th229 100.%-a Th228 100.%-a Ra225 100.%-Jl-Ra224 100.%-a Ac225 100.%-a Rn220 100.%-a Fr221 100.%-a Po216 100.%-a At217 OR At217 99.%-a 1.%-fr" Pb212 100.% - fr" Bi213 OR . . . Bi213 Rn217 97.84%-fr" 2.16%-a 100.%-a Bi212 OR . . . Bi212 64.07%-f, 35.93%-a Po213 Tl209 Po213 100.%-a 100.%-fr" 100.%-a Po212 Tl208 100.%-a 100.%-Jr Pb209 Pb209 Pb209 100.%- fr" 100.%-fr" 100.%- fr" Pb208 Pb208 Bi209 Bi209 Bi209 Fig. 1.12. Decay chains for 2320 and 2330.
1-25 at I m from an unshielded point source is then summarized for the activity in a fuel salt package and a flush salt package. It is seen that the external dose is dominated by 137 Cs. The 208TI daughter from 23 2U decay is also a significant contributor, and because of its higher energy gamma, is more difficult to shield. These values indicate that, without significant shielding, a fuel snit package will exceed 200 mrem/h at the surface and will be classified as remote-handled. However, a flush salt package (possibly with minor shielding beyond the thickness of the container wall) is likely to be a contact-handled package. Table 1.7. Inventor y of actinide isotopcs 0 Isotope Half life Curies Activity in fuel salt Activity in (Ci) flush salt (Ci) mu and its daughter chain 27.4 kg 1JJU 0.2 kg 1JJU mu 70y 129 113 0.81 ll8Th 1.9 y 132 116 0.83 21 Ra 3.66 d 132 116 0.83 110Rn 55.6 s 132 216p0 116 0.83 150 ms 132 116 0.83 lllpb 3.25 h 132 116 0.83 lllBi I.0lh 132 116 0.83 lllpO 45 s 84.8 74.3 0.53 1osr1 3.05 Ill 47.6 41.7 0.30 mu and its primt1ry daughter chain mu J.59 X JOSy 302 265 1.90 ll9Th 7300 y 0.8 0.70 O.QJ llSRa 14.8 d 0.8 0.70 0.01 11sAc JO d 0.8 221 0.70 0.01 Fr 4.9m 0.8 0.70 0.01 ll7At 32 ms 0.8 0.70 0.01 ll3Bj 45.6111 0.8 0.70 0.01 lllpo 4 i1s 0.8 0.70 0.01 l09pb 3.25 h 0.8 0.70 0.01 Other significant uranium and transuranic isotopes 2.5 kg 2J4U 0.02kgm u ll-lU 2.45 JOSy X 17.4 15.2 0.11 98.2%P11 J.8%Pu l38pU 87.7y 0.89 0.87 0.02 ll9PU 24,1 IO y 41.7 40.9 0.75 l Opu 6,540 y 15.3 15.0 0.28 l lpU 14.4 y 212 208 3.82 l 'Am 433 y 23.2 22.8 0.42 Summary q.factinidrt inwlllorii!s Total actinide activity, 931 815 5.9 Z.-::92 Total uranium activity 448 393 2.8 Total transuranic 293 288 5.3 activity Total activity 1,672 1,496 14.0 "Bnsed on decay to December 1999; all uranium pr.:sent except that which hns migrated from fuel salt ~otal inventory of3 l .3 kg 211ll and 2.8 kg "'U. mw=u ratio assumed to be constant
1-26 Table 1.8. Dose at I m for unshielded radionuclides in postulated fuel and flush salt packages Fuel salt Flush salt Isotope Dose at Im for I Activity in Dose at Im Activity in Dose at Ci (R/hr) package (Ci) (Rlhr) package (Ci) Im (R/hr) 137 Cs 0.33 229 76 3.97 1.309 IS2Eu 0.58 0.045 0.026 154 0.0008 0.0005 Eu 0.62 0.13 0.079 0.0022 0.0014 mEu 0.03 0.18 0.005 0.0031 0.000) 232l.J 1.3 4.7 6.1 0.034 0.044 mu 0.029 11.0 0.32 0.079 0.0023 234u 0.078 0.64 0.050 0.0047 0.0004 Notes: Fuel nnd flush salt divided into 24 packages, each; inlcud.:s uranium except that which migrated to off-gas system. Dose values for fission products from Radiological Health Handbook, January 1970, p. 13 I. 137
*& radiation included in 137Cs abow. '°Srl"'Y have no gnn1111as. "Tc has no gaJllllla.
msm has only a weak gamma (0.022 MeV at 4% yield). 147 Pm has no gamma. Values for uranium isotopes from ORNUM0-45 (given a.~ rem'11 in air). Value for 232U includes daughter chain. in equilibrium. Figures 1)3 and 1.14 depict the fission product and the actinide and transuranic activities over time. Cesium and strontium will continue to dominate the fission product activity until several hundred years after discharge, at which time these 30 year half-life isotopes will have decayed away. Then, 93Zr and ~c will dominate the fission and activation product dose into geologic time frames. However, once the 90Sr and 137 Cs isotopes have decayed away, the total fission and activation product activity will be about 10 Ci. Similarly, the 232U decay chain *will dominate the actinide activities until it is overtaken by ingrowth 23 of 3U daughters after a few hundred years. Hundreds to thousands of curies of actinide activity will be present out into geologic time fran1es. Were the uranium separated from the salt at the time it is removed from the tanks, the 233 U chain will not grow in to any significant extent, and the 232 U daughter products would decay with the 1.9 year half-life of 228Th. The long tem1 inventories would then be represented by the plutonium and americium chains, and would remain below 100 Ci after about 50 years. Thus, separation of uranium can have a significant impact on the long-term radionuclide inventories in salt packages. Of course, the inventories associated with uraniwn and its daughters are not eliminated, but rather transferred to a uranium package. 1.2.4 Chemical Characteristics Today, both the fuel and flush salts arc a mixture of LiF, BeF2 , ZrF4 , UF4 , and trace quantities of the fluorides of other corrosion products, fission products, actinides and transuranics. The phase diagram for the system LiF-BeF2 -ZrF 4 is shown in Fig. 1.15; adding UF4 to the system necessitates a three-dimensional representation (plastic models have been used to depict such systems in the past). A series of experiments was perfom1ed to quantify the crystallization sequence of the MSRE salt. Table 1.9 summarizes the crystallization sequence for MSRE fuel salt (Thoma 1971 ).
1-27 ORNL DWG 85-438 10 6 TOTAL FISSION PRODUCTS AND DAUGHTERS 10 5 - - - Ce-144/Pr -144 AND Pm-147
....' \ ...... ...
\
*** * * * * * *
- Sr-90/Y-9 0 AND Cs-137/Ba -137m
~ 10 4
\ ----- Zr-93/Nb- 93m AND Tc-99 f-
\* \
( The Zr-93 is an activation product.)
> 10 3 \ \
f-0
\
C 0 10 2
\
a: ..J i! \ 0 t- 10 1
'\
10° ---------------=-*-:: .. 10-i,...__ _..__ _......______ . _____ ______, 10° 101 10 2 10 3 10 4 105 10 6 TIME AFTER DISCHARGE (years) Fig. 1.13. Fission product activity of MSRE fuel and flush salts.
1-28 ORNL DWG 85-439 10 6 . . . . . . - - - - , - - - - - - - , . - - - - - - . - - - - - , - - - - . - - - - TOTAL ACTINIDES AND DAUGHTERS U-232 CHAIN U-233 CHAIN Pu-239 , Pu-240 , AND - Pu-241 /Am-24 1 10-i ........_ _......______,.._____ ____ ____ ______ 10° 10 1 10 2 103 104 10 5 106 TIME AFTER DISCHARGE (years) Fig. 1.14. Actinide and daughter activity ofMSRE fuel and flush salts.
1-29 ORNL*DWG 66-739lR3 Zr~ 903 PRIMARY PHASE AREAS: TEMPERATURES IN *c @LiF COMPOSITION IN mole % @ Li6 Bef.iZrF8 @ Li 2Be~ @ Li 2 Zrfi. @ Li 3 ZrF7 Li3Zr.iF19 @ BeF2 @ Zrf.t 2-LIQUIOS Fig. 1.15. Phase diagram of the LiF-BeF2-ZrF4 system.
1-30 Table 1.9. Crystallizati on sequence for Molten Salt Reactor Experiment fuel salt Crystalline form Onset temperature Volume fraction (OC) Li 2BeF4 434 0.735 Li 2ZrF6 431 0.204 LiUF5 416 0.032 BeF2 complete -350 0.029 Note: Crystals of pure BeF2 are generally not expected. The last liquid solidifies as a glass which incorporates variable amounts of the phases listed above with BcF2* An experiment was perfom1ed in which simulated fuel salt was frozen at rates that would approximate the cooling rate in the drain tanks after reactor shutdown. The objective of this experiment was to detennine whether uranium segregation could take place as crystalline phases began to fonn but most of the salt was still liquid. Uranium samples taken from the frozen salt plug ranged from 2.94 wt% near the top to 5.74 wt% at the bottom. This degree of segregation was considered small enough to not affect criticality safety in the drain tanks. One key reason these salts were selected for use in MSRE is that there is very little volume change between the solid and liquid phase. It has been estimated5 that, on melting Li2 BeF4 , the density is reduced only about 2.07%. Other physical properties for liquid salt were measured as part of the MSRE development program, and are documented in the literature. Data found in the MSRE literature for the physical properties for solid salt is more sparse, but open literature data for components in the salt can be used along with the data that is available for the MSRE salt to obtain reasonable estimates for the key properties over the necessary range of temperatures. The UF6 and F2 gas that has migrated away from the salt has left the salt in a net reducing condition. It has been estinlated that at least 115 mot of fluorine was generated by radiolysis and removed from the solid (Williams et al. 1996). This represents a net 23 0 equiv of reductant present in the fonn of isolated metal sites (Li 0 and Be 0 ). The present reducing potential of the stored salt is latent in the solid fonn, but once the salt is melted the reducing potential of these sites can be realized, and the metal species will react according to their redox potentials, Li > Be > U - Zr. The following reactions are a consequence of this reduction series: 2Li 0 + BeF2 - 2LiF + Be0 , 3Be0 + 2UF3 - 2U0 + 3BeF2 , and 2Be0 + ZrF4 - Zr0 + 2BeF2
- 1-3 I The reduction of berylliwn by lithium is expected to be kinetically favored and the cascade of subsequent reduction steps probably eventually converts all of the uranium to U3+ and some fraction of the uranium and zirconiwn to the metallic state. By using a lower bound of 23 0 equiv of reductant (based only on UF6 detected in the vapor phase) and an estimate of 142 mol of uranium in the fuel salt, it can be projected that after all UF4 is converted to UF3, 88 equiv of reductant remains. The close proximity of uranium and zirconium makes it difficult to predict the subsequent reduction of these metals. Were the remaining reductant to reduce uranium preferentially, a mass of 6.8 kg of uranium metal would be formed. Other concerns include plating of uranium on the tank walls, and precipitation or disproportionation ofUF3* Thus, adjustment of the redox chemistry of the salt will likely be required before or during melting.
Preliminary experiments have indicated that production of fluorine is the result of beta-gamma, not alpha, radiation. Using the the fission product decay energy and the most reasonable fluorine yield curves measured (Williams et al. 1996), the amount of fluorine that will be generated in the future is depicted in Fig. 1.16. This cun*e is most representative of a case in which uranium is separated from the salt after the salt is removed. If the uranium remains with the salt, production of UF6 must also be considered, and the gamma radiation from the mu and 233 U daughter chains must be taken into account. The present storage of the salts has been evaluated relative to Resource Conservation and Recovery Act (RCRA) regulatory status (Nix I992a,b). It has been concluded that the salts, in their present form, are neither listed nor characteristically RCRA hazardous waste. In particular, Skipper ( 1992) detennined that beryllium fluoride is not a listed waste and, as a solid, does not exhibit the characteristics of ignitability, corrosivity, reactivity or toxicity as defined in the RCRA legislation. This detennination only applies to storage in its present fom1, and must be reviewed in the context of any processing perfonned on the salt. 1.2.5 Summary of All Locations Containing Fuel or Flush Salt Table 1.10 provides a list of all locations in the MSRE that contain fuel or flush salt, along with estimates of the quantity of uraniwn present in the salt. A review of the components likely to contain salt was made, and process history was used to detem1ine the type of salt that was last circulated through that component. The largest quantity of salt outside the fuel and flush (and coolant) drain tanks is a salt heel left in the fuel storage tank after the fuel salt was fluorinated in 1968. When the fuel salt was returned to the drain tank prior to the addition of the 233Uenriching salt, a quantity of salt was left in the fuel processing cell for transfer to a distillation experiment in the spare cell. When salt was transferred to the spare cell, measurements indicated Urnt only a. fraction of the desired amount could actually be transferred. Although a number of reasons were postulated, it was generally assumed that this indicated that a hole had appeared in the dip leg during the corrosive fluorination operation. Based on the available measurements, it was assumed that 190 kg of salt remained in the fuel storage tanks. This salt would contain essentially no uranium (after fluorination), and would contain the fission products that existed at the end of 235 U operations. More recent investigations into the conditions in the fuel processing cell (DeVore et al. I996) did not identify radiation fields that would correspond to the expected cesium inventory. Thus, U1e actual amount of salt in the fuel storage tank may be substantially less.
1-32 Projection of Fluorine Accumulation Based Upo n GF = 0.02 mol ecul es/l0 0eV 2 300 ,--~-,--,---,-,-~-,-,----r--r--r-r--,-,-
-r-r-r---.--.--,
I 250 N ~ 4-c 0
~
-~ 0 r.f.) (1) 200 l I /
/ -- -
...- P--1 /
~ 0 150 s
P--1
=:s u
e I I I u 100 1996 I < Purge and Trap 1 1 50 I I I I 0 I 1 10 100 Years after 1970 Discharge Fig. 1.16. Projected fluorine generation resulting from fission product decay.
1-33 Table 1.10. Summary of Molten Salt Reactor Experiment salt inventory Location Type of salt Mass of salt Mass of (kg) uranium (kg) Fuel drain tanks Fuel salt 4,650 <32.7 Flush drain tank Flush salt 4,290 0.5 Fuel storage tank Fluorinated fuel salt 190 negl. Distillation experiment Fluorinated fuel salt 36 negl. Reactor cell Flush salt 20 negl. Drain tank cell piping Flush or fluorinated fuel salt 12 negl. Release into drain tank cell Fuel salt 0.1 negl. Fuel processing cell piping Fluorinated fuel or fresh salt 9 negl. Coolant drain tank Coolant salt 2,610 none Other estimates of salt volumes were made in a similar manner. Most of the salt in the piping systems was that left in the freeze valves. Drawings for freeze valves were reviewed, and process history was used to determine the type of salt that last flowed through the freeze valve. In all cases, the last salt in the valve was either flush salt or fluorinated fuel salt, and again no valve contains any significant uranium inventory. Salt in the distillation experiment is the same as that in the fuel storage tank, with some separation between salt in the feed and receiver vessels in the experiment. It had been estimated that 20 kg of salt remained in the reactor circuit after the salt was drained into the drain tanks; again, the last salt circulated in the reactor was flush salt. When the failed freeze valve 105 was removed, visual examination was used to estimate that only a few cubic inches of salt were left in the cell. Although this was actual fuel salt, the low quantity again means there is no significant uranium inventory associated with the spill. Thus, none of the known inventories of salt outside the fuel and flush drain tanks contains any significant quantity of uranium. If the inventory in the fuel storage tank is substantially <190 kg, there may be only about 100 kg of salt outside the tanks. The fission product inventory in this quantity of salt may be about 260 Ci of fission products. In performing the CERCLA feasibility study, a review will be made as to whether the selected salt removal technology (assuming the salt is removed) is readily adaptable to removing any of the salt outside the fuel and flush tanks. If so, that salt removal will be recommended for inclusion into the scope of the task. If not, the salt will be left for disposition as part of the overall decommissioning of the facility. 1.3 PRESENT CONDmON OF THE MSRE FACILITY The MSRE experiment was ended in calendar year 1969 so that funding for the following year could be used instead for the design of a prototype Molten Salt Breeder Reactor. Thus, at shutdown, a few key safety signals were tied into a remote monitoring station at the ORNL Waste Operations Center, and the facility operations staff was disbanded. After the radionuclides in the salt had decayed for a year, a small staff was reassembled for post-operation examinations and to implement a number of additional containment measures. The reactor cell was opened, test specimens and pieces of reactor components were removed for examination, a visual inspection of the reactor assembly was made, and the cell was sealed closed again. One of the control rods was left permanently inserted in the reactor assembly.
1-34 A heavy accumulation of dust (about 1/2 in. deep on flat surfaces) was observed inside both the reactor and drain tank cells (Cagle and Pugh 1977). An ordinary water hose was used to flush the dust onto the cell floor. By keeping the cell wetted, no contaminated dust was allowed to escape. At the same time, the drain lank cell was opened for inspection and for removal of failed freeze valve 105 and the nearby freeze valve I 06. Copper plugs were inserted into the ends of the salt lines leading back to the fuel drain tanks, using a mechanical clamping mechanism. A blind flange was inserted between the flange faces of a connection in the off-gas line leading from the tanks to the pump bowl in the reactor cell. The cell was then sealed closed, and has not been opened since. Various penetrations into the cell were cul and capped, including the helium addition lines that had been used to pressurize the tanks for sail transfers. The reactor cell was opened one more time in 1977 to measure radiation levels at the top of the reactor graphite and at various other locations in the cell (Cagle and Pugh 1977). The effective tenth-value layer of lead was detem1ined to be 1.4 in. All MSRE equipment except the items required for surveillance and operation of the drain tank and salt processing systems was placed on a surplus materials list and made available to other projects. Control and electrical panels for U1e drain tank and salt processing systems were eventually painted pink or yellow to make it obvious that equipment was not to be removed. The process computer was transferred to the TVA Bull Run steam plant, and the control rooms were dismantled. The coolant salt pump and piping were removed for use in a molten salt reactor component development test loop. The 5 in. coolant lines were welded shut where they exit the reactor cell. The scope of surveillance and maintenance activities over the years was generally limited to monitoring the overall status of the facility. Checks were made by a waste operator making rounds through that part of ORNL and entered into the log. The annual annealing procedure was performed, through 1989, as dictated by the procedure (Guymon 1971). A reactor operator from the adjacent High Flux Isotope Reactor perfom1ed the annealing operation. Other spaces in tl1e MSRE complex were used to house personnel and equipment from the Health Sciences Research Division al ORNL. The most use was made of the offices in Buildings 7509 and 7503, but areas such as the high bay and open areas on the lower floor of Building 7503 were used for storage as well. The buildings were maintained as needed for this occupancy. At one point, a fence was placed in the high bay to prevent casual entry to the reactor, drain tank, and fuel processing cells. Some required facility maintenance was perfom1ed once the MSRE Remediation Project was formed. Improvements were made to the data collection system that read, among other parameters, the temperatures in tl1e drain tanks and the drain tank cell. Routine trending of data is now taking place. Facility maintenance has included repair of the roof at several locations, including over the electrical equipment in the switch house, repairs to the steam system, and similar activities. The building cranes have been recertified. More significantly, the facility is now staffed with knowledgeable operations, maintenance, and health physics personnel. Portal monitoring and access control has been improved, and the spare cell and fuel processing cell have been accessed for inspection. Some of the techniques used to inspect these cells might be applicable to a preliminary inspection of the drain tank cell. The drain tanks are th~ught to be in good condition. A continuous review of corrosion product concentrations in the salt during reactor operations did not indicated any significant damage to the Hastelloy-N tank and thimble walls. The recent sample taken from the off-gas system did not contain large quantities of volatile fluoride corrosion products. Furnace heaters have been operated annually until
1-35 December 1989, and no operational difficulties have been observed. Equipment for removing shield plugs over the drain tank cell remains at lhe facility. The maintenance shield is in place over one of the other cells, and is used periodically to access irradiated materials specimens stored in the cell by another ORNL division. The condition of the off-gas system, including valves in the system, will be determined as the off-gas system is evacuated by the reactive gas removal system. Once the reactive gas removal system has purged the gases out of the tanks and refilled the system with helium, the tanks themselves can safely be opened. The reactive gas removal system will remain as an operating trapping system for use during salt removal operations. The solidified salt in the tanks has never been inspected. It is thought to be a monolithic mass, although some fracturing may have taken place during cooling. A them1al model of the salt" has replicated the measured temperatures in the drain tanks using thermal conductivity values for solid salt, indicating that gross degradation of the sail block, to the extent that additional thermal resistance is introduced, has not happened. The condition of existing salt processing equipment is less certain than that of the drain tanks. Corrosion rates during the fluorinalion operation conducted in 1968 were significant (Lindauer 1969). Operational difficulties have suggested that a dip leg in the fluorination vessel had corroded through; however, recent inspections into the cell did not identify radiation levels consistent with the projected snit inventory in the fuel storage tank (DeVore et al. 1996). Visual inspections of the cell contents indicated that, at least on the exterior, equipment in the fuel processing and spare cells is in remarkably good condition. There is no dust layer; the impression is that much of the equipment had only recently been installed. Radiation levels in the fuel processing cell are moderate. Radiation levels on the side of the spare cell that contains the sail processing off-gas equipment are quite low, and personnel have entered the cell for further survey work and to replace a filter near the cell floor. No contamination has been observed on the exterior of equipment in either cell. Other systems that might be needed for salt transfer and processing operations have been removed, and would have to be replaced. This includes the helium system used to pressurize the tanks to transfer snit, and the fluorine station used during the fluorination operations in 1968. The migration of uranium to the fuel processing and spare cells indicates that the freeze plugs do not provide a tight seal, although remelting and then freezing the plugs might restore lhe seal. A general survey of the activities needed to reactivate existing drain tank and salt processing systems is underway. General building systems are available to support operations at the drain tanks and at the salt processing system. These include electrical and other utilities, the building ventilation systems, monitoring systems (including criticality and radiation monitors), and fire protection systems. The flowrate on the building ventilation system had been reduced to the requirements of storage and surveillance to lin1it the power consumption and increase the service life of the stack fans. The original flow might have to be restored before shield plugs are removed from the drain tank cell. The building is adequate in tem1S of protecting equipment and workers from weather, and reasonable truck and crane access is available in the high bay. The high bay confinement was constructed as a separate structure inside an existing Building 7503. At present, there are multiple openings in this structure as a result of minor architectural modificnti011S and the removal of equipment for salvage. The ventilation system does maintain airflow into the structure. Repairs to this confinement barrier could improve the negative pressure in the high bay relative to the rest of the building. "DAC-EA-020794-A0 19 N.d. Thermal Analyses ofMSRE Fuel Drain Tanks, in preparation.
1-36 1.4 BASELINE RISK ELEMENTS A streamlined risk assessment is being prepared to evaluate the potential human health risks associated with the fuel and flush salts at the MSRE, assuming no actions are taken to remove the salts or mitigate contaminant concentrations or exposures (personal communication from P. Howell to F. J. Peretz, Oak Ridge National Laboratol)', July 24, 1996). This assessment is being prepared to the guidance generally applied to CERCLA feasibility studies for an interim remedial action. Most probable pathways to exposure have been identified for the near tem1 and the long tem1. In the near tem1, the most probable exposure pathway is the continued production ofF and UF , 2 6 collection of these gases in the off-gas system, and a sudden release due to a failure of a compone nt in the off-gas system. This risk exists today (until operation of the reactive gas removal system begins). An approximate assessment of the consequences of such an event can be obtained by reviewing the system safety analysis for the reactive gas removal operation (SSA/7503-ERP/003/R0 1996). Accordin g to a scenario in which a component of the now pressurized off-gas system suddenly fails and releases the entire gas inventOI)' in the off-gas system, 50% of the gas is retained in the building, the rest is lost as a ground level release, and typical conservative atmospheric conditions are assumed. A dose commitment of 24 rem is projected for a member of the public at the nearest accessible location on Bethel Valley Road just outside of the main ORNL complex. Over time, the generation rate of fluorine (and probably UF6) will decline. This pathway will be of far less concern once the primal)' gamma source, 137Cs/137mBa, has decayed away in a few hundred years. For the long term (hundreds to thousands of years), the exposure pathway begins with water intrusion into the drain Lank cell or the drain tanks themselves. Building 7503 was originally construct ed in the early 1950s for use by the aircraft reactor project, and was modified several times as it was adapted to new uses. In 1957, construction of an Aircraft Reactor Test reactor was stopped as part of the downscaling of the overall aircraft reactor program. In the intervening years until construction of the MSRE began, the building was largely abandoned. At that time, it contained negligible quantities of radioactive material. The present reactor cell existed, but the drain tank cell had yet to be construct ed. The basic system of french drains around the reactor cell and the sump and pumps southeast of the reactor cells was in place. Thus, groundwater at the site has been suppressed by pumping for about 40 years. During this intervening period, a failure of the sump pumps went undetected for several weeks. Groundwater rose to an elevation about chest high in the sump room (personal communication from T. C. Morelock to F. J. Peretz, Oak Ridge National Laboratol)', July 24, 1996). This water level correlates to an elevation of 823 ft above sea level, or approximately the elevation of the large opening between the reactor and drain tank cells as shown in Fig. 1.17. Other reviews of water level in the area indicate that a likely undisturbed year-round average groundwater elevation would be about 835 ft above sea level (personal communication from P. Howell to F. J. Peretz, Oak Ridge National Laboratory. July 24, I996). Even at 823 ft, the fuel and flush drain tanks arc totally submerged, as is much of the salt and off-gas system piping above the tanks. At present, these components are well protected from groundw ater because of the operation of the sump pwnps and the physical barriers provided by the concrete cell walls, the stainless steel liner, and the integrity of the tanks and piping themselves. In the long tenn, however, it cannot be assumed that pumping will continue, and eventually the barriers can fail. The failure need not be Lo the cell walls and liner directly, but could be a line that penetrated the cells below the water level. It is very possible that the solidified sail in the drain tanks has fractured as it cooled. This provides a mechanism for waler enll)' into the salt itself, once water enters the tank interior. A simple calculation has shown that if a 2% void in the salt is filled with water, a kerr of 1.06 is obtained (Crume I994). A criticality in a fuel salt drain tank then leads to exposure pathways ranging from direct exposure to radiation to the generation and dispersal of gaseous fission products.
ORNL OWG. 14*591 2 3 4 5 6 7 8 9 1 I J AND IO*JON CIIANU r--- ---- -:, STACk 1 I ..
, -.:,~ :..:,- 71 I I
I CONJIIOL I1
=-==--=--=-~J11 110011 L.
- l'
-I l,J
-...J 1
LIQUID WAST[ C[LL '.\ i
'1 Elevation 823 ft ...:::7 ,uu PIIOCUSIN G C[LL A[ACfOA V[SS[L fH[AIIAL SHl[LO PU[L DIIAIN UN* flO I Fig. 1.17. Appro ximate elevation to which water has risen in sump room with sump pumps out of operat ion.
1-38 This mechanism is absolutely prevented when the dominant fissionable isotope, 233 U, decays away with a 1.6 x 105 year half-life. Although the barriers to ,:vater intmsion are very effective in the short term, reliance on such barriers for over a hundred thousand years is not credible. Other secondary pathways have also been identified. These include the gradual dissolution of the salt once groundwater enters the drain tanks, and transport of a range of contaminants to nearby receptors. These pathways are generally slow, chronic processes, and pose a lesser risk to the potential receptor than the gas release or criticality scenarios. 1.5 OUTLINE OF THE ALTERNATIVES EVALUATION PROCESS The remainder of this report directly addresses tl1e identification and evaluation of alternatives for the disposition of the MSRE fuel and flush salts. In Chap. 2, potential end points for the key contaminants in the salt are identified, and the key factors that address the potential and requirements for placing the MSRE material in each end point are discussed. In Chap. 3, the technologies available for removing the salt from the tanks, for separating uranium from the other key contaminants, and for stabilizing materials as waste or product are identified and discussed. In Chap. 4, the removal, separations, and stabilization technologies are used to develop real scenarios, in many cases using existing facilities for performing the necessary operations to place the key contaminants in the identified end points. Of these scenarios, a representative sampling is identified as the list of alternatives for evaluation. In Chap. 5, an evaluation matrix is prepared for these alternatives, leading to a set of recommendations. Chap. 6 identifies the key uncertainties that could affect these evaluations, and the activities that are recommended lo resolve the uncertainties. The overall logic for identifying and evaluating alternatives is shown in Fig. 1.18. The evaluation begins with the salt in its present conditions, assuming no other actions are taken to mitigate the transport of the key contan1inants to potential receptors. The present condition of the fuel and flush salts is evaluated in the context of the baseline risk established in the previous section. Because unacceptable consequences result fro a no action" alternative, this condition is not selected as the ultimate end point, and further evaluation is necessary. The next step evaluates pern1anent storage in the drain tanks, with additional features in place to prevent the transport of key contaminants. Because the criticality issue continues for hundreds of thousands of years the alternative of long-tem1 storage is deemed unacceptable and the evaluation proceeds on the basis of the removal of the salts .. Once removed, the salts are evaluated against the requirements for disposal or reuse against the full set of potential end points, as shown in Fig. 1.19. If the salt in its present condition cannot meet requirements for any of these end points, further activities are taken. In the logic depicted in Fig. 1.18, the next step is to consider separating the fissile and long-lived actinide activity hazards from the rest of the salt. This results in two streams, each of which is again evaluated against the requirements of the end points. If the requirements are not satisfied, further stabilization steps are considered. Once material meets requirements for disposal, the process slops. If no disposal path based on an accepted end point is identified, the result is long-term interim storage of a unique waste form. In the evaluation matrix. the identified alternatives arc evaluated against the seven CERCLA evaluation criteria. In addition, two other criteria are used. Because the requirements for many of the potential end points, such as the federal repository, are either not fully developed or are subject to change as a result of external factors. the alternatives are evaluated in terms of the ease in which a waste form
Salt, as-is Interim storage Stop Disposal Stop activities Enhance Stabilize Interim I storage uranium storage i
- iI J
--1 Stop Disposal Stop i activities
-w
\0 Remove for Separate Stabilize disposal uranium salt I
I I .~ Disposal Disposal Disposal activities activities activities Stop Stop Stop Fig. 1.18. Logic diagram of the alternatives identification process.
\
I Can it be disposed of? e Transfer Disposal Disposal Disposal
- I
.i::,.
0 1 for reuse I I activities I I activities I I activities I I Disposal activities ( Stop ) { Stop ) { Stop ) { Stop ) l Stop Fig. 1.19. Expanded logic of the disposal evaluation block. '*1 I l
1-41 or product specification can be changed to address the potential new requirement. Adaptablilty to changing requirements is added as an evaluation criteria to address this issue. Second, there are objectives for the fuel salt disposition task that are not necessarily addressed by the CERCLA process. These include a conunitment to fulfill the DNFSB 94-1 commitment that the salt will be removed and the uranium will be placed in storage as an oxide, and a preference to remove the salts (or clearly establish a permanent disposition at the site) so that the final decommissioning of the MSRE systems in Building 7503 can proceed. Consistency with programmatic objectives is added as an evaluation criteria to address this issue. The full set of criteria used, beginning with the seven CERCLA criteria, is:
- Overall protection of human health and the environment
- Compliance with applicable or relevant and appropriate requirements
- Long-term effectiveness and pcnnancnce
- Reduction of toxicity, mobility, or volume through treatment
- Short-tern1 effectiveness
- Implementability
- Cost
- Adaptability to changes in system requirements
- Compatibility with programmatic objectives The primary evaluation of alternatives is perfonned using scenarios judged to be preferred, or at least representative, to achieve the desired end point. Once a recommendation is made, the selection of scenario for that alternative is reviewed in more detail. Again, all credible scenarios that achieve the recommended end point are listed in an evaluation matrix, and an evaluation at this second-tier level is conducted to refine the recommendation. This matrix again uses the CERCLA criteria, with any additional evaluation criteria judged appropriate added to the matrix.
2-1
- 2. POTENTIAL END POINTS 2.1 MSRE DRAIN TANK CELL The most obvious potential end point for the MSRE fuel and flush salts is to remain stored in the drain tanks in the drain tank cell until all radionuclides have decayed to innocuous levels. The drain tanks and the drain tank cell have been described earlier. Further modifications could include additional sealing of penetrations, the cutting and capping of all off-gas lines, and sealing the penetration between the reactor and drain tank cells. The drain tank cell (or even the tanks themselves) could be filled with grout or some other material. Neutron absorbers could be placed in the tanks, on top of the salt, or around the tanks in the drain tank cell. Likewise, the salt could be stored in its present condition or it could be removed, treated in a number of ways, and returned to the tanks.
The drain tank cell, a heavily reinforced concrete structure with a stainless steel containment barrier, is a substantial structure and is located below grade. It is, however, at a lower elevation than the natural water table. It is located at a key intersection in the Melton Valley portion of ORNL, and part of an attractive site for future utilization. As part of the ORNL site, it is a controlled industrial site. The MSRE site is near a number of other radioactive waste facilities, including solid waste burial sites, liquid waste seepage pits, and the hydrofracture waste disposal facilities. The pennnnent disposal of radioactive materials in at least some of these facilities in Melton Valley is assured. There is presently no defined re&*tdatoiy basis for pernmnent disposal of the salts in the MSRE drain tank cell. This issue would have to be addressed in the Federal Facilities Agreement site treatment plan, and would have to be agreed to by TDEC and EPA Region IV. Disposal of materials classified as transuranic waste (transuranic activities greater than 100 nCi/g) in shallow land facilities may be prohibited by DOE Orders. If permanent disposal in the drain tank cell were to be allowed at all, the requirements for disposal would likely be performance based. A detailed risk assessment would be performed to demonstrate that the risks to human health and the environment near the MSRE site would be acceptably small. 2.2 REUSE Reuse of the snit or components from the snit can be described in a number of ways. The entire mass of salt may have reuse potential in another molten salt reactor or as a target or blanket material in a neutron-producing accelerator system. In some neutron-producing accelerators, the target incorporates a subcritical assembly, making it similar to a reactor core. Salt may be useful for an actual pilot or production facility, or may be used in the research and development activities surrounding the development of a reactor or accelerator project. The latter includes research into the separation of fuel, target, or other radioactive materials from a reactor or accelerator system. Other reuse end points could apply to the fissile 233 U separated from the snit, or to all or components of the carrier snit after separations. The lithium in the snit is isotopically pure 7Li. In some cases, the key contaminants in the salt may be accepted in the reuse option. In others, the contaminants must be removed first, and n nonradioactive or otherwise separated material would be transferred to the new user. In the case of the latter, another end point must be found for the contaminants separated from the salt before its transfer.
2-2 The attractiveness of the reuse end point is a matter of whether or not a credible use option exists. Although no significant molten salt reactor development activities have been identified, an active program to develop accelerator-driven neutron sources for transmutation of plutonium or other materials in nuclear waste is rn1derway at Los Alamos National Laboratory. This program is currently funded at a relatively low level, but does support laboratory work for the design and development of pilot and production level facilities. Recent contacts have been established with Los Alamos personnel (Toth 1996). These contacts have led to a request by the Director of the Los Alamos Neutron Scattering Center and Energy Research Programs to explore the details of a transfer of salts from the MSRE to Los Alamos (Brown 1996). The most immediate interest is in the nonradioactive coolant salt. Requirements for storage, handling, and the ultimate use of the flush and fuel salts will also be explored. Storage and use of these salts may require a higher level offunding than is presently available. Even in the near term, however, the transfer of small quantities of flush or fuel salt may be useful for the laboratory development of salt processing systems. No programs to develop reactors that use 233 U as fuel are active in the United States today. Use in a solid-fuel reactor (traditional reactor designs that use fuel rods or plates) requires fuel fabrication facilities that are designed to handle the radiation associated \vith the uranium isotopes present. Uranium with a high 232U content is particularly rn1likely to be desirable as a reactor fuel. In general, the individual components of the salt do not appear to be particularly valuable. A possible exception is high-purity 'Li. Other systems, including tritium production systems, have required significant quantities of 6Li. The residues from 6Li processing could provide a significant source of 7Li, although the isotopic purity of this material has not been checked. As chemicals, the individual components of the salt are readily and inexpensively available from commercial sources. From a regulatory standpoint, it is more desirable to reuse an existing contaminated material than place it in a disposal facility and contaminate another material. Specific regulations for the handling and transportation of radioactive materials would have to be met in the course of transferring salt or components from the salt to a new user. Physical requirements for the material to be transferred would be established on a case-by-case basis, balancing economics against the needs of the new user. Material requirements could include some level of decontamination, chemical purity, and packaging. 2.3 WASTE ISOLATION PILOT PLANT 2.3.1 General Description The Waste Isolation Pilot Plant (WIPP) has been authorized, designed and constructed to provide for the safe disposal of defense-related transuranic waste, including waste generated during the cleanup of nuclear weapons production facilities. It is part of an overall transuranic waste program that provides an integrated management system from generation through transportation and disposal. Disposal of commercial, high-level waste, or spent nuclear fuel at WIPP is not allowed (DOE Carlsbad Area Office 1995; DOE/CAO 95-1095 1995). WIPP is located in southeastern New Mexico, 26 miles east of Carlsbad, New Mexico (Fig. 2.1). Surface facilities include a waste handling building, designed for receiving and inspecting waste containers and preparing the containers for transfer underground. The disposal areas are excavated into the 25 million year old Salado bedded salt fomrntion 2,150 feet below the earth's surface (Fig. 2.2).
lronment mology S RFETS) N Ji w Savannah River Sile (SAS) Waste Isolation Pilot Plant (WIPP) i Highway Legend
- j Interstate Highways ~
I
! U. S. Highways I
Fig. 2.1. Location e>f the Waste Isolation Pilot Plant, indicating transportation routes from major DOE sites. (" ' .. ,.
2-4
-surr*1s 1c1a an d F eet I IL __ Ground Level Dewey Lake Redbeds Mudston e and Siltstone 540 -
Rustler Formation lnterbe dded Layers 850 1000 Salado Formation 2000 2159 Waste Repository Level Evapo rites (Salt) 3000 Sea Level 3400 Sea Level Castle Formation Salt and Anhydrite
--- 4000 -
Bell Canyon Formation 4500 Fig. 2.2. Geologic formations at the Waste Isolation Pilot Plant.
2-5 The waste disposal area within WIPP consists of eight panels, each containing seven rooms between access drifts. Waste is classified into contact-handled transuranic (CH-TRU) and remote-handled transuranic (RH-TRU), depending on whether the dose at the surface of the waste package exceeds 200 mrem/h. The RH-TRU wastes will be emplaced in a manner different from that for the CH-TRU wastes because of packaging, shielding, and loading requirements; operational equipment; and structural considerations. The disposal configuration of the CH-TRU waste inventory calls for emplacement of the waste packages in disposal rooms that measure 33 ft wide, 13 ft high, and 300 ft long (Fig. 2.3). The configuration for RH-TRU disposal provides for emplacement into the walls of the disposal rooms in horizontal boreholes. These boreholes will be drilled 4 feet from the floor on 8-foot centers. The CH-TRU waste inventory will be emplaced in the disposal rooms following completion of all RH-TRU waste disposal activities in that room. A total of ~7,955 RH-TRU waste boreholes will be provided in WIPP. The CH-TRU waste packaging will include 55-gallon drums, standard waste boxes, and ten drum overpacks. The standard waste box is an oblong steel box that is 37 in. high, 54.25 in. wide, and 71 in. long. The ten drum overpack is a welded-steel cylinder 74 in. in dian1eter and 74 in. high. CH-TRU waste containers will be transported and received in TRUPACT-IJ shipping containers. These containers are eight feet in diameter and l Ofeet high, and are provided with inner and outer stainless steel containment vessels. They satisfy U.S. Department of Transportation standards and are certified by the Nuclear Regulatory Commission. Conventional diesel tractors will pull trailers designed to transport as many as three TRUP ACT-II containers. RH-TRU wastes will be received at WIPP in Nuclear Regulatory Commission (NRC)-certified type B shipping containers. One canister, which holds ~3 55-gallon drum equivalents of waste, will be placed into each borehole. Each RH-TRU canister is made of 0.25-inch-thick carbon steel, is 121 in. long and 26 in. in diameter. A general depiction of the RH-TRU canister is given in Fig. 2.4. A shield plug will be inserted into the borehole after the canister is emplaced to protect workers from radiation. The RH-TRU canisters are to be transported in the RH-TRU 72-B shipping cask. All waste containers, both CH-TRU and RH-TRU, will be vented to prevent the accumulation of gas in the container. Revision 5 of the WIPP Waste Acceptance Criteria (DOE/WIPP-069 1996) has recently been issued. The purpose of the document is to provide transuranic waste generator and storage sites with the minimum requirements transuranic waste must meet before transport to and disposal at the WIPP. Revision 5 combines in a single document the criteria from several sources, including the WIPP Safety Analysis Report, the TRUPACT-II Safety Analysis Report for Packaging, the Resource Conservation and Recovery Act Permit Application, the WIPP No Migration Variance Petition, and the WIPP Land Withdrawal Act. The WIPP Transuranic Waste Baseline Inventory Report (WTWBIR) data gives inventory quantities and waste material parameters for TRU waste. These data, as well as the radionuclide inventory, are supplied by the DOE waste generator/storage sites. Summaries of the radionuclide and non-radionuclide inventories obtained from the WTWBIR will be used in the WIPP performance assessment process. Generator/storage sites' anticipated remote-handled waste inventories are likely to fill the WIPP remote-handled waste capacity of 250,000 cubic feet. Most of the stored remote-handled transuranic waste will comefrom ORNL. Total allowed repository capacity is 6.2 million cubic feet.
ROOM CROSS-SECTION 1~ 300* I 13' t e 0 e 0 e e 0 0 a e e 4@.-ll.li1 I 1 4( CH-TRU Waste Drums N I 0\ ster and s PLANVIEW Not to Scale Fig. 2.3. Arrangement of the remote handled transuranic waste disposal facilities at WIPP.
?i I i-------------------121*-------------------i i I
--j f--1.1*
Lti1 0.25" 26" 25.5" 6.6" __l_ 9.1" N I
-..J 5.1" Fig. 2.4. Dimensions of a remote handled transuranic waste canister.
2-8 All transuranic waste destined for Lhe WIPP will be certified for disposal. The generator sites will make sure that all transuranic waste meets waste acceptance criteria before being shipped to the WIPP site. The WIPP will oversee this certificalion process to assure that generator site certifications are being carried out to acceptable standards. After the facility has reached its capacily, the underground openings will be backfilled with salt, the swface facilities will be removed, and seals will be installed in drifts and shafts. The plastic, self-healing nature of the salt formation will also contribute to the sealing process. WIPP is now entering a waste handling demonstration mode in preparation for disposal operations. In October 1997, after completing all activities in the WIPP Disposal Decision Plan, including receiving all necessary permits and certifications, Lhe Secretary of Energy is expected to notify Congress of the decision to operate the WIPP as a disposal facility. After a mandatory six-month waiting period, the first container of transuranic waste is scheduled to arrive al the site in April 1998. Initial operations will be with CH-TRU only. RH-TRU operalions are to begin in fiscal year 2002. 2.3.2 Regulatory Basis Construction of WIPP was originally authorized by the U.S. Department of Energy National Security and Military Applications ofNuclear Energy Authorization Act of 1980 (Pub. L. 96-164, 93 Stat. 1259). This act defines the mission of WIPP as a facility for the disposal of defense-related transuranic waste. More recently, the WJPP Land Withdrawal Act (Pub. L. I 02-579, I 06 Stat. 4777) provides for the transfer of the land on which WIPP is constructed to the Department of Energy, and outlines the activities required for bringing WIPP into operation. As a result of this legislation, various agencies are given certain responsibilities for promulgaling regulations, preparing and reviewing documents, and certifying compliance with regulations. The DOE will complete construction, evaluate readiness, and bring WIPP into operation under its system of Orders and Price-Anderson legislation. DOE will prepare a Supplemental Environmental Impact Statement (EIS) for Disposal Operations before the declaration that WIPP is ready to operate. An Implementation Plan for this supplemental EIS was issued in May 1996. The DOE will also maintain and implement a Disposal Decision Plan, and the Secretary of Energy will notify Congress that WIPP is ready to operate when all actions in that plan have been completed and all certificates of compliance have been obtained. There is a mandatory 180 day waiting period following this notification before waste disposal operations can begin. The EPA has two major roles in the certification of WIPP. EPA must issue a No-Migration Determination to allow the disposal of mixed rndioaclive and hazardous transuranic waste at WIPP. This will be issued in response to a No-Migration Variance Petition that has just been submitted to EPA by DOE/CAO in July. The second role held by EPA is to define the specific criteria to be used for demonstration of WIPP compliance with Lhe technical standards for radioactive waste disposal provided in the Code of Federal Regulations (40 CFR 191), and to accept a Final Compliance Certification Application that will be submitted by DOE. The EPA recently finalized the specific criteria for the WIPP compliance demonstration. The final disposal criteria, promulgated as 40 CFR Part 194, was approved by the EPA Administrator on February I, 1996. A companion document, referred to as the Compliance Application Guidance, will further interpret the final criteria. DOE expects to submit the Compliance Certification Application in October, I996. Completing the necessary activities to apply for compliance certification is a substantial task. but DOE/CAO intends to meet the schedule.
2-9 The New Mexico Environment Department is responsible for issuing a RCRA Part B pennit for operation ofWIPP as a mixed-hazardous waste disposal facility. DOE submitted a pennit application to New Mexico in May 1995, and has recently responded to fomml comments on it. The Disposal Decision Plan calls for the pem1it to be issued in August 1996. The NRC issues licenses for transportation of waste to WIPP. Requirements for transportation are reflected in Revision 5 of the waste acceptance criteria. The NRC recertifies the transportation containers on a regular basis. The TRUPACT-11 was recertified in August 1994. The RH-TRU 72-B cask will be certified when WIPP is closer to operation with RH-TRU waste. The Land Withdrawal Act specifies a number of other required activities before WIPP can open. These include completion of the RH-TRU study (DOE/CAO 95-1095 1995), a review by the National Academy of Sciences, and various notifications of Congress, states, and Indian tribes. 2.3.3 Disposal of MSRE Salts in WIPP WIPP is an attractive end point for the key contaminants in the MSRE salts for a number of reasons. It is the only repositoiy with a reasonable probability of opening in a time frame consistent with the fuel salt disposition task schedule. As a repositoiy, it provides for true ultimate disposition of the salt contaminants. The operating mode, waste package, and waste acceptance criteria are defined. Confinement ofradionuclides is provided by the repositoiy salt matrix, and extensive processing of the MSRE salt may not be required. A proposed package for the fuel and flush sail residues is depicted in Fig. 2.5. An inner Hastelloy-N can is filled with molten salt. This can is constrncted of 0.5 in. thick walls to provide an adequate corrosion allowance should uranium be separated from the salt by fluorination in the can. The can is then inserted into a shielded container with the approximate outer dimensions of a 55 gallon drum. This container provides two inches of steel in each direction as a radiation shield. There are a total of 24 cans each of fuel and flush salt, packaged three to an RH-TRU canister as shown in Fig. 2.6. Thus, there are 8 RH-TRU canisters for the fuel salt, and 8 for the flush salt. Each salt can is filled with salt to - 75% ofits internal volume. The rest of the volume (at least, in the fuel salt packages) accommodates a soda-lime getter package. The can and shield structure will be vented through the getter and HEP A filters, as required by the WIPP packaging requirements. Uranium will be separated from the fuel salt to remain within the WIPP limit for fissile material. This will also greatly reduce the long-tenn radionuclide inventoiy in the salt by eliminating the 232U and 23 3{] decay chains. Table 2.1 shows the total inventoiy of radionuclides in the salt in 200, 500, and 1000 years (DeVore 1996a). Separation of the uranium affects the radionuclide inventoiy in two ways. It speeds the decay of the 232 U chain by removing the 72 year half-life 232 U parent, leaving the chain to decay with the 1.9 year half-life of 228Th. It also prevents the grow-in of the 233 U daughter chain. It is the growth of that chain that causes the radionuclide inventoiy to begin to grow between 500 and 1000 years in the case where uranium is left with the salt.
2-10 f--------------22.25*----------i
~Steel
/
V- Hastelloy-N Waste
-- 27.75 1 32.75 1 f--0.50 1 Storage Volume 2.00*
17.25 1 Fig. 2.5. Proposed shielded container for MSRE salt.
Package 1 Package 2 Package 3 NI Fig. 2.6. Depiction of three MSRE salt packages in an RH-TRU waste canister.
2-12 Table 2.1. Radionuclide inventory in all of the fuel and flush salt sent to the repository (Ci) Years after disch~trge Inventory in Inventory in salt fluorinated salt" containing all uranium 200 years 355 879 500 years 80 518 1000 years 68 603 "Based on residual uranium concentration of20 ppm in both fuel and flush salt Radiation dose estimates were also made for both the fuel and flush salt (DeVore 1996b), based on the geometry of Fig. 2.5. These estimates are shown in Table 2.2. Table 2.2. Dose rates at side surface of postulated fuel and flush salt containers Shielded Unshielded 0 Fuel salt residue IOR/hr 270 R/hr Flush salt residue 0.36 R/hr 4.5 R/hr "Measured at the outer surface of the outer container, with a 2 in. gap where the shielding would have been For the purposes of comparing this package to the WIPP waste acceptance criteria, it is assumed that the uranium will be removed by fluorination of molten salt, as was done at the MSRE in 1968. The residual levels of uranium were 26 ppm in the fuel salt and 7 ppm in the flush salt. Fissile and other radionuclides that would be present in the fuel and flush salt residues were tabulated, and the appropriate comparisons to the quantitative limits expressed in Revision 5 of the WIPP waste acceptance criteria are shown in Table 2.3. It is seen that in all cases, the MSRE salt residues fall within the limits of the waste acceptance criteria. It is also seen that both salt residues clearly exceed the limits for defining material as transuranic waste.
2-13 Table 2.3. Comparison offuel and flush snit parameters in a remote-handled transuranic canister to Waste Isolation Pilot Plant (WIPP) waste acceptance critierin° WIPP limit Fuel salt Flush salt Volume of salt in canister 0.235 1113 0.24m3 Mass of salt in canister 581 kg 533 kg Specific activity" 23 Ci/L 13.4 Ci/L 0.22 Ci/L 23 9Pu equivalent activity 1000 PE-Ci 36PE-Ci 0.05 PE-Ci 23 9Pu fissile gram equivalents~ 600 g package 95 g 4g 325 g cask Surface dose rated 100 rem/hr IOR/hr 0.36 R/hr (no more than 5% at higher dose) 1000 rem/hr Thermal power 300 watts < 14.2 watts <0.25 watts Transuranic content > 100 nCi/g 61,900 nCi/g 1,240 nCi/g 0 Assumes the fuel salt is placed in eight RH-TRU canisters, each containing three containers. 6 Assumes 98.3% of all activity is in the fuel sail (ratio given in ORNLffM-13142 for fission products). decayed to December 1999. "Based on the sum of 233 U, mu. and 239Pu inventories in fluorinated sail. "calculations perfonned at contact with the side of a sail can with 2 in. of steel for shielding. Unshielded surface doses are 270 R/hfor the fuel sail and 4.5 R/hfor the flush salt. 'Thermal power not reduced to account for uranium migration or fluorination. The long-tenn impact of this material on the overall perfom1ance of WIPP would seem to be trivial. There are a total of 16 RH-TRU containers (assuming that shielded flush salt is not regarded as contact-handled waste). The total radionuclide inventol)' after 500 years (a time for which the substantial containers in which the salt is packaged can be assumed to maintain integrity) is only 80 Ci, and the chemical form of the waste is that of a halide salt, similar to the material of which the repositol)' is constructed. Additional work is required to ensure that the addition of fluorides does not in some way significantly change assumptions for container lifetimes in the repositories. The long lifetime that should be achieved by the MSRE containers, constructed of materials compatible with the fluoride salt, does make it likely that fluorides would not be released until other containers are assumed to have lost their integrity. The issues that do exist with disposal of the MSRE salt residues are programmatic, not technical. The WIPP Land Witl1drawal Act, and ilie earlier legislation auiliorizing construction of WIPP, designate WIPP for the disposal of defense transuranic waste. The first major issue is whether the MSRE salt residues are transuranic waste, or if they are spent nuclear fuel or high level waste. Some definitions of these three waste classifications are provided as Table 2.4. Sec. 12 of the WIPP Land Withdrawal Act states: The Secretal)' shall not transport high-level radioactive waste or spent nuclear fuel to WIPP or emplace or dispose of such waste or fuel at WIPP.
2-14 Table 2.-t. Definitions of waste classifications High Level Waste Definitio~s DOE Order 5820.2A, Attachment 2, Definition #18: The highly radioactive waste material that results from the reprocessing of spent nuclear fuel, including liquid waste produced directly in reprocessing and any solid waste that contains a combination of transuranic waste and fission products in concentrations requiring permanent isolation. NRC 10 CFR 60.2: High-level radioactive waste orHLWmeans: (l) Irradiated reactor fuel, (2) liquid wastes resulting from the operation of the first cycle solvent extraction system, or equivalent, and the concentrated wastes from subsequent extraction cycles, or equivalent, in a facility for reprocessing irradiated reactor fuel, and (3) solids into which such liquid wastes have been converted. NRC 10 CFR 60.2: Isolation means inhibiting the transport of radioactive material so that amounts and concentrations of this material entering the accessible environment will be kept within prescribed limits. NRC 10 CFR 60.2: Pen11ane111 closure means final backfilling of the underground facility and the sealing of shafts and boreholes. EPA,40 CFR 191.02(h): High-level radioactive waste, as used in this part, means high-level radioactive waste as defined in the Nuclear Waste Policy Act of 1982 (Pub. L. 97-425) Nuclear Waste Policy Act of 1982 (Pub. L. 97-425), Section 2, (12): The term high-level radioactive waste means (A) the highly radioactive material resulting from the reprocessing of spent nuclear fuel, including liquid waste produced directly in reprocessing and any solid material derived from such liquid waste that contains fission products in sufficient concentrations, and (B) other highly radioactive material that the Commission, consistent with existing law, determines by mle to require pennanent isolation.
- Department of Energy Spent Nuclear Fuel Programmatic Environmental Impact Statement Implementation Plan: High-Level Waste - I) !JTadiated reactor fuel, 2) liquid waste resulting from the operation ofthe:first cycle solvent e>-.1raction system, or equivalent, and the concentrated waste from subsequent extraction cycles, or equivalent, in a facility for reprocessing iJTadiated reactor fuel, and 3) solids into which such liquid wastes have been converted.
Spent Nuclear Fuel DOE Order 5660.1, Definition (4.q): In*adiated nuclear material that contains fission products, requires shielded storage and handling facilities, and is in an unusable fonn. DOE Order 5820.2A, Attachment 2, Definition #33: Fuel that has been withdrawn from a nuclear reactor following irradiation, but that has not been reprocessed to remove its constituent elements. NRC, 10 CFR 72.3: Spent Nuclear Fuel or Spent Fuel means fuel that has been withdrawn from a nuclear reactor following irradiation, has undergone at least one year's decay since being used as a source of energy in a power reactor, and has not been chemically separated into its constituent elements by reprocessing. Spent fuel includes the special nuclear material, byproduct material, source material, and other radioactive materials associated with foe! assemblies. EPA, 40 CFR 191.02 (g): Spent nuclear foe! means fuel that has been withdrawn from a nuclear reactor, the constituent elements of which have not been separated by reprocessing.
2-15 Table 2.-l (continued) Nuclear Waste Policy Act of 1982 (Pub. L. 97-425), Section 2, (23): The tenn spent nuclearJuel means fuel that has been withdrawn from a nuclear reactor following iffadiation, the constituent elements of which have not been separated by reprocessing. Department of Energy Spent Nuclear Fuel Programmatic Environmental Impact Statement Implementatio n Plan: Spelll Nuclear Fuel (SNF) Material used as the fuel for a nuclear reactor that has undergone nuclear fission, including mateiial such as leftover um*eacted fuel, fission products, transuranic elements, and various fuel assembly hardware. Spent fuel can be reprocessed to recycle the unused portion of the fuel and to separate the other components. Transuranic Radionuclides or Waste DOE Order 5820.2A, Attachment 2, Definition #38: Transuranium Radionuclide. Any radionuclide having an atomic number greater than 92. DOE Order 5820.2A, Attachment 2, Definition #38: Transuranic Waste. Without regard to source or fonn, waste that is contaminated with alpha-emitting transuranium radionuclides with half-lives greater than 20 years and concentrations greater than I00 nCi/g at the time of assay. Heads of Field Elements can detennine that other alpha contaminated wastes, peculiar to a specific site, must be managed as transuranic waste. EPA,40 CFR 191.02 (I): Transuranic radioactive waste, as used in this part, means waste containing more than 100 nanocuries of alpha-emitting b*ru1surruuc isotopes, with half-lives greater thru1 twenty years, per gram of waste, except for: (I) High-level radioactive wastes; (2) wastes that the Department has detennined, with the concurrence of the AdmiJustrator, do not need the degree of isolation required by this part; or (3) wastes that the Commission has approved for disposal on a case-by-case basis in accordance with IO CFR Part 61. Sect. 2(12) of the Land Withdrawal Act provides the follov.ring defmition of high-level radioactive waste: The tenn "high-level radioactive waste" has the meaning given such tenn in section 2(12) of the Nuclear Waste Policy Act of 1982 (42 u.s.c. 10101(12)). The Nuclear Waste Policy Act of l 982 (42 U.S.~. _970 I) defines high-level waste: The tenn "high-level radioactive waste" means (A) the highly radioactive material resulting from the reprocessing of spent nuclear fuel. including liquid waste produced directly in reprocessing and any solid material derived from such liquid waste that contains fission products in sufficient concentrations; and (B) other highly radioactive material that the Commission, consistent with existing law, detennines by rule to require pennanent isolation. The MSRE fuel and flush salts are presently identified in the DOE Integrated Spent Nuclear Fuel Database, and the MSRE fuel salt is briefly addressed in the Programmatic Spent Nuclear Fuel Management and Idaho National Engineering Laboratory Environmental Restoration and Waste Management Programs EIS (DOE 1995). Separation of uranium from the salts, although not for the purpose of reuse, might be construed as reprocessing. The tem1 "highly radioactive material" is not
2-16 quantitatively defined. The dose rates at the surface of an unshielded container (as shown in Table 2.2) are not inconsequential. However, the activity concentration in the fuel salt residue is 13 Ci/L, for less than in the reference DWPF borosilicate glass waste (Baxter 1983). The Savannah River waste contains 177,000 Ci in 22.1 ft3 (205 L), or about 860 Ci/L. The radiation level at the surface of a reference Savannah River waste canister is 5500 R/h. Although not presented as a threshold for high-level waste, the Land Withdrawal Act does place a quantitative limit of 23 Ci/L for waste to be emplaced in WIPP. A draft recommendation on distinguishing RH-TRU from high level waste (Watkins 1996) was recently presented to the National TRU Steering Committee. This recommendation proposes two decision points. The first is whether the waste is irradiated fuel elements, or reprocessed, or liquid or solid wastes resulting from reprocessing of, irradiated fuel elements. The draft guidance recommends classification of these wastes as high level waste. The second decision point is whether samples of fuel elements were irradiated only for R&D purposes and not to produce power. If so, it is recommended that they be considered for disposal in WIPP. The recommendation goes on to state that the document should not be interpreted that fuel used to "power" test reactors be clarified as RH-TRU, but rather be classified and managed as spent nuclear fuel. The second issue is whether the MSRE sail residues are defense related, or commercial. Although an outgrowth of defense programs to develop molten salt aircraft reactors, the MSRE was research perfom1ed by tl1e Atomic Energy Commission directed at the development of a commercial reactor. MSRE itself was in no way a commercial facility, but a government research facility. Definitions of defense that have included all research activities at DOE sites that perform a defense mission have been proposed. A recent memorandum from the DOE General Council (Nordhaus 1996) would narrow the definition of defense dramatically. The general council recommendation is to interpret defense as only wastes derived from the manufacture of nuclear weapons and operation of naval reactors, and associated activities such as the research in weapons laboratories. Wastes generated by DOE's environmental management program in its cleanup and management of weapons production waste is also construed as acceptable for disposal at WIPP. Before the MSRE salt residues can be disposed of at WIPP, these issues must be resolved. The salt residues appear to meet all quantitative requirements for disposal at WIPP, but may clash with policies, including those expressed in the Land Withdrawal Act. It may be necessary to obtain a favorable legal finding, or even await future legislative activity, before the MSRE salts can be accepted at WIPP. Aside from resolution of policy issues, other activities will be necessary before disposal of MSRE salt residues in WIPP can be approved. As yet, the actual certification process has not been initiated for the MSRE salt residues. The MSRE salts do not appear in the WIPP Transuran"ic Waste Baseline Inventory Report. These activities will be initiated if it is detem1ined that there is a reasonable chance that the MSRE salts could be accepted in WIPP. 2.4 FEDERAL REPOSITORY 2.4.1 General Description The Nuclear Waste Policy Act of 1982 (42 U.S.C. 9701), along with other legislation, authorizes activities leading up to the siting, construction, and operation of a geologic repository for the disposal of high-level radioactive waste and spent nuclear fuel. The federal repository is to be operated by the Department of Energy, under a license issued by the Nuclear Regulatory Commission that ensures that standards established by the EPA, with the guidance of the National Academy of Sciences, are met. The
2-17 Nuclear Waste Policy Act establishes a Nuclear Waste Fund, with payments made by generators and owners of spent nuclear fuel and high-level wastes to fund the repository and associated transportation and waste acceptance programs. A presidential decision in 1985 directed that the repository also be used for the disposal of radioactive wastes resulting from atomic energy defense activities. A General Fund appropriation is made to offset the costs associated with the disposal of defense-related wastes. Three sites were nominated out of an initial evaluation of 9 sites; this evaluation has currently been narrowed by law to the Yucca Mountain site in southern Nevada, about 66 miles northwest of Las Vegas and adjacent to the Nevada Test Site. The Yucca Mountain site has not been selected as the repository site; it is only the focus of an intense characterization activity. The characterization activities are part of an overall program for the management of such wastes managed by the DOE Office of Civilian Radioactive Waste Management. For the purposes of demonstrating that the Yucca Mountain site is capable of supporting a repository, a conceptual design of repository facilities is being proposed. An overall depiction of that design is shown in Fig. 2.7. A ramp extends down and into the mountain from a north portal, passes through the waste emplacement zone, and circles back to a south portal. Upper and lower waste emplacement blocks are accessed from that tunnel. As part of the characterization activities now underway, the tunnel is being bored from the north portal into the waste emplacement areas. As of March 1996, over 2.5 miles of tunnel had been drilled, and various alcoves arc being completed and fitted out with experiment equipment. Completion of a thcnnal testing alcove, that will be fitted out with experiments that simulate waste-rock interactions with wastes generating significant heat, and an alcove that accesses the Ghost Dance Fault, allowing studies as to how that geologic structure may influence water movement, is expected to be completed this year. The budget request for 1996 was not fully provided ($400 million was provided as opposed to a requested $630 million and a 1995 level of $512 million. Of the funds provided, $85 million was reserved for development of an interim storage facility that would begin accepting spent fuel from utilities, when such a facility is authorized. Since that facility has not yet been authorized, the effective funding provided in 1996 is $315 million. As a result, a new program plan is being prepared, either changing the scope or date of earlier commitments. The plan is now to produce a viability assessment of the Yucca Mountain site in 1998, as opposed to a full characterization report. Contingent on approval of a more streamlined licensing process, an application to the NRC for construction permit in 2002 may still be possible. The goal is to begin emplacement of nuclear waste in the repository in 20 I 0. Development of a multi-purpose canister for the transport, storage, and disposal of spent fuel elements from commercial reactors had been underway but has been affected by the budget restrictions. The Navy is supporting limited development of the concept for transportation of naval fuel to Idaho National Engineering Laboratory (INEL). A notice of intent to begin preparing an EIS for a federal repository at the Yucca Mountain site was issued in 1995. This work, too, has slopped because of a lack offunding. 2.4.2 Regulatory Basis The regulatory basis for the activities to develop a federal repository is provided by the Nuclear Waste Policy Act of 1982, as nn1ended. Major amendments, focusing the characterization effort on only the Yucca Mountain site, were enacted in 1987. Further amendments are before Congress now. The proposed amendments include direction for tl1e immediate construction of spent-fuel interim storage facility so that DOE can begin to accept fuel from contributors to the Nuclear Waste Fund in 1998. If
GH08TDANCE FAULT ,1
-I
'.1 J I N I 00 UPPER l!MPLACl!Ml!N T BLOCK Fig. 2.7. Depiction of the federal waste repository under consideration at the Yucca Mountain site.
2-19 enacted, this direction may distract atlenlion from direct progress on the Yucca Mountain characterization and subsequent repository design and licensing activities. In 1992, the Comprehensive National Policy Act specified that unique standards were to be established for the protection of public health and safety from releases of radioactive materials stored or disposed of in the repository at the Yucca Mountain site. The standards are to be the only such standards applicable to the Yucca Mountain site; notwithstanding other standards authorized by the Nuclear Waste Policy Act or the Atomic Energy Act. Within 90 days of enactment, the EPA was to contract with the National Academy of Sciences to conduct a study to provide findings and recommendations on reasonable standards for protection of the public health and safety. This study was issued in final form in 1995 (Anon. 1995). Within ayearofreceiving the recommendations of the National Academy of Sciences, the EPA is to promulgate public health and safety standards for protection of the public from releases from radioactive materials stored or disposed ofin the repository at the Yucca Mountain site. These standards are to be the only such standards applicable to the Yucca Mountain site. Within a year of when the EPA promulgates its standards, the Nuclear Regulatory Commission is to modify its technical requirements and criteria as given in l OCFR 60 to be consistent with the EPA standards. A number of review boards are mandated by the legislation guiding the development of the repository. Among these is the Nuclear Waste Technical Review Board, with membership nominated by the National Academy of Sciences. The Nuclear Waste Policy Act prohibits U1e emplacement of spent fuel containing more than 70,000 metric tons of heavy metal or an equivalent quantity of high-level waste resulting from the reprocessing of such a quantity of spent fuel until a second repository enters operation. 2.4.3 Disposal of MSRE Material in the Federal Repository The federal repository provides for the ultimate disposal of spent fuel and high-level waste. If the MSRE salts, salt residues, or material separated from the snits are defined as spent fuel or high-level waste, the federal repository is the only proposed facility that can ultimately accept the wastes. Although characterization of the Yucca Mountain site is well underway, selection of that or any other site for the repository is in the future, and operation of a federal repository is even further in the future. Program plans call for operation of the repository in 20 IO at the earliest. The approach to setting standards and licensing the repository is complex, involving the DOE, EPA, NRC, the National Academy of Sciences, and a nwnber of 0U1er advisory boards. The process is expensive, and it is difficult for the government to fund it. Issues such as U1e immediate construction of an interim spent fuel storage facility can further distract from progress on the repository. Two other issues cloud U1e likelihood of MSRE wastes entering the initial federal repository. As noted, the capacity of U1e initial repository has been limited by law until a second repository opens. Wastes in that amount have already been identified; in a sense the initial repository is already fully subscribed. Second, the repository is funded by commercial utilities and contributions from the DOE defense program. Again, the MSRE wastes are neither the result of a commercial activity or, under the strictest definitions, the result of defense activities.
2-20 Standards developed to date for the repository address the overall performance of the repository. To date, no waste form has been licensed or qualified for disposal in the federal repository, although a large body of knowledge has been accumulated on borosilicate glass, which is the leading candidate waste form for high-level waste and is being produced at DWPF and the West Valley Demonstration site. Repository performance assessments performed to date consider retention of radionuclides in the waste form itself (unlike the WIPP perfonnance assessment, which is based on the characteristics of the host salt to seal the waste into a dry environment). Thus, the salts in their present form, which is slightly soluble, are not likely to be acceptable in the federal repository. High-level waste is managed at only a few large sites, including Savannah River, West Valley, Hanford, and INEL. ORNL does not have any waste classified as high-level waste, or any facilities for handling such waste. Such facilities are generally expensive. Qualification of a waste fonn for disposal in the repository is also expensive. It may be that the only credible path to the repository would be to, in some way, incorporate the MSRE material into one of the waste forms produced by a site with a significant high-level waste program. 2.5 FISSILE MATERIALS DISPOSITION PROGRAM The fissile materials disposition program has been fonned to address the disposition of fissile materials that are no longer needed for their mission, especially materials stockpiled for weapons systems that will not be produced or material from weapons systems that are being dismantled with the cessation ofthe nuclear arms race between the United Stales and the former Soviet Union. The primary focus of this program is the disposition of excess plutonium. However. the disposition of highly enriched uraniwn, including the relatively small (compared to the more common 235 U or 239 Pu isotopes) amount of 233U that has been accumulated, is also being addressed by the materials disposition program. Currently, the bulk of the U.S. inventory of 233 U is stored al ORNL and at INEL. The material stored at ORNL is mostly in the fom1 of oxides, whereas most of the 233 U inventory at INEL is in the form of fuel elements. The 233U stored at ORNL is kept in the Radiochemical Development Facility (Building 3019), shown in Fig. 2.8. About 500 kg of 233U is in storage, mostly as U 0 or other oxides. 3 8 A significant fraction of this is the result of the Consolidated Edison Uranium Solidification Project (CEUSP), which them1ally converted a uranium nitrate solution to oxide. The CEUSP 233U has a similar 23 20 content to the MSRE uranium. The remainder of the 233U has a low 232U content. It was used in the Light Water Breeder Reactor prototype core installed in the Shippingport reactor. The uranium oxide is placed in welded canisters that are 3.5 in. outside diameter and 24.25 in. long. The canisters are then placed in a can that is 3.625 in. outside diameter and 24.75 in. long. The canisters, in their protective cans, are stored in wells, many of which are drilled into the concrete walls between the cells of the RDF as shown in Fig. 2.9. Additional storage wells remain available for future use. There is presentlY, no use identified for the CEUSP 233 U inventories in storage. As yet, they have not been fonnally declared surplus. The low mu material may have a use as a source df9 Th for medical applications. The inventories arc presently managed by DOE Defense Programs. The DOE Materials Disposition Program has requested a program plan from ORNL for the disposition of the 233 U inventories. This has resulted in the investigation of disposition alternatives in many ways similar to this study (Kocher N.d.). Disposition options that have been identified include incorporation into a glass matrix (possibly by blending into the DWPF feed stream) with ultimate disposal in WJPP or the federal repository. Fission or transmutation in a reactor or accelerator are other options, although the cost is difficult to justify for the relatively low overall inventory of 233 U.
3017 CHEMICAL TECHNOLOGY LABORATORY CJ 3123 I F.n __ __ __ 0 3020 I 3074 3138 3100 I SHOP 3001 OPERATIONS BUILDING \ (GRAPHITE REACTOR) 8 DOCK 2025 ANALYTICAL
~
a HOT ANALYTICAL 3019 N LABORATORY I IX FACILITY N
~ 3019B rc:r~s]r L_____r,1rcp1
__ L_'!rc-p.1r~
..JL __ JL __1rcc:n JL_J 1str,tt£ orTRUST 3130 3025 WASTE OPERATIONS SOLID STATE DIVISION 2010 CONTROL CENTER LABORATORY 3037 ORNL SURFACE SCIENCES CAFETERIA LABORATORY Fig. 2.8. Site plan for the Radiochemical Development Facility (Building 3019) at ORNL.
ORNL 0-Q 70*S612R u .. ........... ,,.. ~ Ott-Gll~I
.... ,H .. lcl, l
- P 4 .....
Lt** ..... r ** I J ..... 0 J I t! N I N N
, *** !I
- '1 *,
HOU: All **mt111lo111 in ,ntflll tac1,1 01 no*** Fig. 2.9. Storage wells for 233 U between cells II and III of the ORNL Radiochemical Development Facility.
2-23 Uranium will be collected from the MSRE off-gas system and auxilimy charcoal bed before the removal of the fuel and flush salts. The uranium in the off-gas system is collected as UF6, and retained initially on NaF traps. A request has been made to allow the temporal)' storage of these traps in Building 30 I 9. This request has led to a request by Defense Programs to transfer Building 30 I 9 and its entire inventory to the Environmental Management Program for eventual stabilization and Iong-tenn storage (Reis N.d.). This would result in declaring the 233 U inventory surplus. As part of the request to use Building 3019 for temporary storage, a commitment has been made to convert the Uf6 that is chemisorbed on the NaF traps to UP 8 within three years of being placed in storage. Thus, theMSRE project is planning a facility to convert UF6 to U3O8 . This conversion facility will be available to the fuel salt disposition task by the time the salts are removed. As noted earlier, separation of the uranium from the salt results in a salt residue that can meet the quantitative requirements of the WIPP waste acceptance criteria. It reduces the Iong-tenn radionuclide inventory (after a thousand years) in the residues to the tens of curies. It also precludes the generation ofUF6, and the potential migration of uranium, in the near tenn while the fission product inventory is still relatively high. Separating the uranium from the salt, converting it to oxide, and placing it in storage in the RDF means that the MSRE uranium will no longer be a unique waste fonn. Funding that will define the disposition of the existing 233 U inventories will address disposition of the MSRE inventory as well. Separation and disposition of the MSRE uranium separately can also serve the long tenn management strategy. The :?33u in the MSRE salts represents most of the long-tenn radiological hazard, as seen in Fig. 1.14. The growth of the radionuclide inventory after a few hundred years, up to about 600 Ci in I 000 years, is due to the ingrowth of daughter isotopes of 233 U and its immediate daughter product, 229fh. The activity of this chain peaks at well over a thousand curies, and does not begin to fall off until about 50,000 years. At these time fran1es, package containers have likely deteriorated. Separation of this material from the bulk of the salt and conversion into a stable oxide results in a material more compatible with the very stable waste fonns, such as borosilicate glass, proposed for the federal repository. Likewise, as the large inventOI)' of fission products decays, the external radiation dose from the waste falls off. Placing the entire inventory of the salts (whether as a salt or as an alternate waste fonn) in a repository creates a deposit of enriched uranium that could be mined. If the uranium is separated by fluorination, it is collected as UF6 , a vapor. At this stage, it could readily be blended with depleted UF6 to reduce the enrichment with isotopic mixing (as opposed to particulate mixing). This downblended UF6 would then be converted to an oxide (and. if desired, incorporated into glass) in a truly proliferation-resistant fonn. 2.6 LOW-LEVEL WASTE STORAGE FACILITIES Low-level radioactive waste is generally stored or disposed of in near-surface facilities. ORNL stores low-level waste on pads, in t1111mlus structures, and in storage wells. depending on the source and radioactivity of the material. Historically, ORNL has operated a number of shallow land burial sites for the disposal of low-level waste. Other DOE sites have similar storage and disposal facilities. Commercial shallow land disposal sites include the Envirocare site in Utah (which only handles low activity material), the Barnwell site in South Carolina, and others. An effort has been underway for a number of years to establish state compacts that would site low-level waste facilities, restricting the use
2-24 of the facility to states in the compact. This process has been controversial. and has generally not led to the opening of new facilities. DOE Order 5820.2A provides the following definition of low-level radioactive waste: Waste that contains radioactivity and is not classified as high-level waste, transuranic waste, or spent nuclear fuel, or l le(2) byproduct material as defined by this Order. Test specimens of fissionable material irradiated for research and development only, and not for the production of power or plutonium, may be classified as low-level waste, provided the concentration of transuranic is less than I00 nCi/g. DOE Order 5820.2A defines transuranic waste as: Without regard to source or fonn, waste that is contaminated with alpha-emitting transuranium radionuclides with half-lives greater than 20 years and concentrations greater than 100 nCi/g at the time of assay. Heads of Field Elements can detennine that other alpha contaminated waste, peculiar to a specific site, must be managed as transuranic waste. Since both the fuel and flush salts contain transuranics in concentrations above I 00 nCi/g, neither .can be considered as low-level waste by characteristic. A classification system for low-level waste has been promulgated by the NRC in IO CFR 61. The highest class of waste considered generally acceptable for disposal in near-surface facilities is called Class C. Limits on Class C waste given in IO CFR 61.55 state that the concentration of alpha-emitting transuranic nuclides with half life greater than 5 years must be less than I 00 nCi/g, and that the concentration of mes must not exceed 4600 Cilln . Again, both the fuel and flush salts contain transuranics in concentrations above I 00 nCi lg. The fuel salt contains 5505 Ci of mes in 1.88 m3, or 2,930 Ci/m3* This major source of external radiation does not exceed the amount allowed in Class C low-level waste; however, when the 6557 Ci of 90Sr is considered against the limit of7000 Ci/m3, the sum of the fission product concentrations in the fuel salt is 113% of the limit for Class C waste (based on decay to the year 2000). The flush salt contains 95 Ci in 1.92 m3, or 49 Ci/m3 of mes, and 113 Ci, or 59 Ci/m3 of 90Sr. The fission product concentration alone does not preclude classification of the flush salt as a Class C low-level waste. Since the salts, as-is or without uranium, are clearly classified as transuranic waste and not low-level waste, disposal of the existing salts as low-level waste is not an option. However, process residues may be evaluated for disposal as low-level waste. It may be necessary that the NRC detennines these residues to be "incidental wastes" to avoid other classifications based on origin. 2.7 INTERIM STORAGE Unlike the previous options in this chapter, interim storage docs not provide for the ultimate disposition of the key contaminants in the MSRE fuel and flush salts. Instead, it provides for the safe storage of the material until another disposition option becomes available. In the extreme (or, for certain short-lived isotopes), storage W1til radionuclides have decayed can become a strategy toward disposition.
2-25 Because interim storage does not place the MSRE materials in a facility for which a perfonnance assessment has been used to demonstrate that risks to the public and the environment are acceptable, and because there are no fomml acceptance criteria for material held in interim storage, an assessment of the hazards associated with storage must be made for each individual case to ensure protection of workers, the public, and the environment. Interim storage could take place at ORNL or at another site that processes the salts or materials from the salts. If stored at ORNL, the MSRE material would likely be stored as salt. This is currently the case in the drain tanks in the MSRE drain tank cell. Hazards associated with the storage of solid salt include the potential for criticality due to uranium in the salt, direct radiation from radionuclides in the salt, the consequences of the potential migration of uranium, uranium daughters, and fluorine as a result of radiolysis, and the consequence of the migration of radionuclides and berylliw11 as a result of other mechanisms. Uranium that migrates as radiolytic gas could result in a criticality hazard in another location, such as currently exists in the auxiliary charcoal bed. Dispersal of radionuclides as a radiol)1ic gas can lead to dose commitments as a result of inhalation. Water can provide a pathway for the transport of radionuclides and beryllium to potential receptors. These issues are discussed for the present storage of the salt in the streamlined risk assessment. Several of the activities proposed as steps in the disposition process can also result in the interim storage of material that poses less risk to workers, the public, and the environment. Separation of uranium from the salt and conversion of that uranium into a stable oxide eliminates the potential for transmission of uranium and uranhun decay daughters as a radiol)1ic gas. Separation of the fission products from the salt, especially cesium, eliminates most of the gamma energy that produces radiolytic gas. If both cesium and the uraniwn decay chains are eliminated, radiol)1ic gas production is likely to essentially cease. Storage of salt at ORNL can take place in the drain tanks, or in a number of other facilities. If fluorinated salt is classified as transuranic waste, it could be stored in the remote-handled transuranic waste facilities at ORNL. Even if noi classified as such, a decision could be made to manage it along with RH-TRU waste. Salt placed in storage must, of course, be placed in appropriate containers that are clearly and properly labeled. The salt in its present fonn is not construed to be a RCRA waste (Nix 1992a,b; Skipper 1992). If materials from the salt are to be stored in other physical fonns, it must be determined whether that fonn is a listed or characteristic RCRA waste. The programmatic EIS for DOE spent fuel (DOE 1995) identifies regionalization by fuel type as the preferred alternative. Fuel type is classified into aluminum-clad fuel. and other fuel. Aluminum-clad fuel is to be managed at the Savannah River Site. Other DOE spent fuel is to be managed at INEL, except for N-reactor fuel which is to be managed at the Hanford site. Since the MSRE salts are obviously not clad with alwninum, the progran1matic EIS designates management at INEL. A study was perfonned by INEL to detennine what storage options would exist for the MSRE fuel and flush salts (Denney 1996). Shipments of spent fuel to the State ofldaho have been the subject of extensive litigation, resulting in a settlement agreement (Settlement agreement N.d.) that places strict limitations on such transfers. The principal conclusions drawn in the INEL study arc that (I) the MSRE spent fuel is projected to be shipped to INEL for storage; (2) all spent nuclear fuel stored at INEL, including MSRE spent fuel if received at INEL, must be removed to a location outside the State of Idaho by January I, 2035. If materials from the salt are treated or reprocessed at INEL, the resulting high-level waste must be ready
_/ 2-26 for shipping by Januruy 1, 2035; and (3) acceptance at INEL is contingent upon the availability or prior construction and preparation of appropriate facilities for receiving the materials. A special concern was expressed relative to the flush salt. This salt is currently listed in the DOE national spent fuel database. However, only the fuel salt is mentioned in the programmatic EIS. If a classification other than spent fuel is determined by the Idaho regulators during the permitting process, there would be no legal mandate for INEL to accept them for reasons other than treatment. The INEL study includes a survey of existing and proposed facilities in which the salts could be stored. Facilities considered include the CPP-749 underground fuel storage vaults, the CPP-651 Unirradiated Fuel Storage Facility, the CPP-603 Irradiated Fuel Storage Facility, the Intermediate Level TRU Storage Facility at the Radioactive Waste Management Complex, and a proposed aboveground dry storage vault facility based on commercially available diy spent fuel storage casks such as the NUHOMS system. Of these, the proposed d1y cask facility is identified as the most credible storage location. The study notes that stabilization of the fuel salt with a getter or by some other processing will be necessary before transferring the salt to INEL. It also identifies the need for further studies to detem1ine shielding, criticality safety, and handling requirements, and observes that adequate funding should be identified for the interim storage as well as for the final disposition of the salts. Interim storage of material in the salts could also take place at other DOE sites as a result of processing in a facility located at that site. Examples would be if the salt were incorporated into glass in the DWPF and the glass were placed in interim storage at Savannah River, if the salts were calcined and stored in the calcine bins at INEL, or processed by electrorefining and stored with similar wastes at Argonne National Laboratory-West (ANL-W). Again, time restrictions on storage in Idaho after processing are established by the settlement agreement. Interstate agreements may also be needed for storage in other states.
3-1
- 3. TECHNOLOGIES 3.1 REMOVAL TECHNOLOGIES 3.1.1 Features of the Drain Tanks As described in the beginning of this report, the solidified salt is now stored in three tanks (two containing fuel salt and one containing flush salt) located in the drain tank cell, at the lowest location in the MSRE building with the exception of the building sump. The tanks are constructed of1/2 in. thick Hastelloy-N, and are suspended from above in a furnace assembly. A computer model of the drain tank system has been assembled, and is depicted in Figs. 3.1 and 3.2. The first figure shows some of the details of the fuel and flush salt drain tank, furnace, and support assemblies. The main difference between the two tanks is the cooling system that is provided for the fuel salt drain tanks. This system includes 32 thimbles that penetrate the head of the drain tank (along with one additional instrumentation thimble), into which an array of bayonet tubes attached to a steam dome was inserted. The thimbles are constructed of 11/2 in. Schedule 40 pipe, which typically has a wall thickness of 0.145 in. The tanks hang from a support ring, which in tum is suspended from weigh cells attached to a system of supports. Each tank has a 3 in. access flange at the top. In the case of the fuel salt drain tanks, this flange is on top of a section of pipe that extends through an opening in the steam dome. Thus, the flange can be accessed without removing the steam dome and bayonet assembly. However, that assembly is itself designed so that it can readily be removed.
The arrangement of the three tanks in the drain tank cell is depicted in Fig. 3.2. The flush salt tank is located closest to the reactor cell. Both bottom shield blocks, resting on support beams, and upper shield blocks, spanning the cell walls, are shown in the figure. The cell is lined with stainless steel, and a stainless steel pan between the two layers of shield blocks completes the sealed containment boundary. A maintenance shield used to access the cells was described at the beginning of this report (Blumberg and Hise 1968). In addition to the mechanical model, a thermal model of the tanks has been developed.* This model has been used to replicate the steady-state temperatures measured on the tank walls and in the instrument thimble, and to track the temperature profile associated with an annealing procedure. These models are is now available as design aids to develop hardware and processes for salt removal. 3.1.2 Removal as Molten Salt 3.1.2.1 Chemistry issues During the period of time (from 1971-present), fission product radiolysis of the solid fuel has generated fluorine through a process represented in simplified fashion (for the major components) as "DAC-EA-020794-A0 19 N.d. Thermal Analyses oJMSRE Fuel Drain Tanks, in preparation.
CIGH CELL
*SALT COOL IIIG SYSTEM
*SUPPORT RIIO OOLIIG TUIIES w
~
UEL DIAII TANK HEATEI Fig. 3.1. Components of the fuel and flush tank system.
*TOP SHIELD IILOCKS
.-vOTTON SHIELD IILOCKS CUL LINER w
w \' COOLING SYSTC FU[L DRAIN TANK II Fig. 3.2. Arrangemen t of the fuel and flush tanks in the drain tank cell, Including cell shielding blocks.
3-4 In addition, fluorine atoms have apparently reacted with and oxidized the UF4 fuel component: In spite of the annual annealing procedure which was conducted from 1970 to 1989 and which was intended to reverse the above radiolytically-induced reactions, significant amounts ofUF and F have 6 2 left the salt. The net result of this migration of oxidized species from the fuel salt is that the salt itself is left in a highly reduced state with a large potential for further reaction in a melt, equivalent in magnitude to the loss of oxidized material. It has been estimated that at least 115 mol of fluorine was generated by radiolysis and removed from the solid (Williams et al. 1996). This represents a net 230 equiv of reductant present in the form of isolated metal sites (Li 0 and Be 0 ). The present reducing potential of the stored salt is latent in the solid form, but once the salt is melted the reducing potential of these sites can be realized, and the metal species will react according to their redox potentials, Li > Be> U ~ Zr, as shown in Fig. 3.3 (Williams et al. N.d.). This figure is based on extensive complication of Baes (1969), and represents the best thermodynamic information available on these salt mixtures. Concentration effects are accow1ted for by the Nernst equation and do not include any corrections for solution phase nonideality at high concentrations. Only elements present in significant quantities (Li, Be, Zr, and U) are considered The following reactions are a consequence of this reduction series: 2Li 0 + BeF2 - 2LiF + Be0 , Li + UF4 - UF3 + LiF, 0 3Be0 + 2UF3 - 2U0 + 3BeF2 , and Predictions of the equilibrium state of the reduced fuel salt when it is melted can be made based upon the values of Fig. 3.3 and our best estimate of the reduced condition of the fuel salt (~230 equivalents ofLi 0 ). There is sufficient reductant to convert all of the uranium to U(III) and some fraction of the zirconium and uranium to the metal. If the thern10dynamic data in Fig. 3.3 were fully representative of the system, it would indicate the formation of zirconium metal, with uranium remaining as UF3* However, the projected reduction potentials of uranium and zirconium are very close, and small errors in estimates, or the effects of previously unknown species or interactions could act to increase the probability of forming uranium metal. The presence of strong reducing sites can be envisioned as a large reduction overpotential that is capable of driving a number of reduction reactions even though they are not favored in the eventual equilibriwn distribution. The potential for wall interactions must also be taken into account (Baes 1969). The consequences of a high ratio of UF/lJF4 are also being investigated (Baes 1969). These include the potential to plate out uranium and zirconiwn on the metal walls of the drain tank or the thimbles, and the possibility of forming a high melting heel. During the l 960s, a series of experiments were performed at the Materials Test Reactor (MTR) at INEL to study interactions between molten salt and the metal alloys and graphite materials proposed for construction of the MSRE. It was during this sequence of experiments that the formation of radiolytic fluorine was discovered. In 1963, a test capsule referred to as MTR-47-5, capsule 36, was irradiated to a fluorine removal of about 2.1%. (This compares to a
3-5 ll.- 50mV for 5% Zr, 0.1 % U at500°C 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 Reduction Potential (Volts) Notes: Potentials referenced to HF/H2 in 0.67 LiF - 0.33 BeF2* All values are for standard electrode potential except for U(IV)/U and Zr(IV)/Zr. These values are corrected by the Nemst equation. Source: Baes, C. F., "The Chemistry and Thermodynamics of Molten Salt Reactor Fuels," Nucl.Me tal.15, 617-44 (1969) Fig. 3.3. Reduction potentials for metals of interest in the MSRE drain tanks.
3-6 fluorine removal of at least 0.13% in the MSRE salts today, with higher removal possible if additional deposits of solid UF6 exist beyond those that have already been identified.) After irradiation, cooling, and measurement of the gaseous product from radiolysis, a melt-out recovel)' of the contents of capsule 35 was conducted to help establish the chemical condition of the radiolyzed salt. This operation was first performed on an unirradiated test capsule and worked flawlessly. However, when the contents of capsule 36 were heated and removed, only about half the salt was removed and the recovered fraction exhibited a 34% reduction in uranium concentration and a 50% reduction in the expected reducing power. It was concluded that substantial reduction of uranium (and possibly other species) must have occurred before the removal operation, and that this reduction produced a high-melting heel that would not drain. It was also noted that the pale green color of the salt drained from capsule 36 indicated that most of the uranium in the melted fraction was present as UF4
- Unfortunately, no analysis of the high melting heel was reported.
In the earlier days of the development of molten salt reactors for the Aircraft Nuclear Propulsion Program, the use ofNaF-ZrF4 mixtures and various alkali metal fluorides (including LiF-NaF-KF, or FLINAK) was explored as a solvent for UF4
- During flow-loop corrosion experiments various reductants (usually ZrH2) were added to the salt to suppress the redox potential for corrosion of chromium in the Inconel piping. These additions were effective in preventing corrosion of chromium, but they also caused reduction of dissolved Zr(JV) and U(IV) to the metal. The addition of ZrH 2 to NaF-ZrF4 -UF4 (56-46-4 mol %) caused two distinct changes: (a) loss of uranium from the sail during reductant addition in the 700°C (1292°F) makeup vessel, and (b) deposition of metallic zirconium (and perhaps some uranium) on the inner wall of the Inconel flow loop. Reduced alkali metal fluorides of various strengths were prepared by controlling the concentration of trivalent uranium during the salt production process. During circulation salts with more than 2% of the uranium as U(lll) a metallic deposit was formed that was judged to be uranium by its metallographic appearance. Although these tests were not performed with the same salt mixture as was used at MSRE, they do demonstrate that the removal of uranium from reduced salt onto metal walls is an issue that must be addressed before the MSRE fuel salt is again melted.
Recently, laboratory tests of salt samples have confinned that the melting of irradiated salts is vel)' different from that of unirradiated salts. The former produces a metallic-like precipitate whereas the latter melts unifonnly to a homogeneous, clear light-green liquid. Further confinnatol)' tests to characterize these precipitates from the irradiated salts are in progress. 3.1.2.2 Melting without chemistry adjustment It would be a simple matter to melt tl1e salts in the drain tanks were chemist!)' adjustment not required. As of the end of the annealing operations in 1989, all of tl1e heaters in the drain tank furnaces appeared to be operating properly. All power and control systems for the heaters have been maintained intact, and (with tl1e exception of the heaters themselves) are easily accessed for inspection or replacement. Even the heaters inside the tank furnaces were designed for remote replacement. Thus, the salts could be melted by simply turning on the heaters. The issue with turning on the heaters is prevention of the potential consequences discussed in the previous section. These include the formation of phases which do not melt at the desired temperature, and the separation of metallic uranium from the salt (although a criticality as a result of precipitation of uranium metal from salt is unlikely). If the uranium plates out on the tank walls then it is not removed from the tank, and a more difficult process for removing the uraniun1 from the tank walls could be required.
3-7 It is possible that the consequences of melting the salt in a reduced condition are reversible. If the snits were first melted, and then sparged with an appropriate oxidant (e.g., HF or F2), the snits might be returned to their original equilibrium. A rigorous demonstration of this procedure with truly representative surrogate salt (and possibly with a sample of actual MSRE fuel salt) would be essential before this strategy could be implemented. Without unequivocal results from such experiments, this approach is not recommended. 3.1.2.3 Adjusting the redox potential before melting The consequences of melting a reduced fuel salt could be avoided if the redox potential of the snit could be adjusted in the solid phase, before melting. This could take place by exposing solid salt to fluorine or HF at ambient temperatures. Since gas analysis indicates that there is 50% fluorine in the off-gas system at the present time, this would appear to be the present condition of the salt, and under this condition fluorine migrates away from the salt. The salt could be exposed to oxidant at elevated temperatures. Fluorine would create more corrosive conditions in the tanks than HF/H2* The oxidant could be carried in an inert diluent gas, such as argon, to achieve a nonexplosive mixture. For this approach to be relied upon, it must be demonstrated that the fluorine diffuses through the solid snit matrix and reaches all of the active metal sites. It must also be demonstrated that the back reaction to the metal fluoride actually goes to completion at the chosen temperature. Previous studies into these reactions sought to establish the temperature at which the recombination back-reaction just counterbalances the generation of fluorine by radiolysis. This is the inhibition temperature. Recombination of fluorine with reduced metal sites was measured by noting the rate of fluorine pressure loss over reduced fuel at various temperatures. Comparison of this back-reaction rate with the radiolysis rate gave reasonably good predictions of the inhibition temperature, at about 150°C (302°F). These predictions agreed with the observations made during and after in-pile irradiation of fuel, in which only moderate heating of the fuel [to above 100°C(212 °F)] was required to suppress fluorine evolution. The primary importance of recombination in the context of melting the salt is in assuring that the radiation damage accumulated in the reduced fuel salt is completely annealed. The ability to inhibit further fluorine evolution by heating is well established. The general issue of annealing radiolytic damage encompasses more variables than inhibition predictions and requires a more thorough understanding of governing phenomena. All of the previous studies using radiolyzed salt have measured recombination rates only during the earliest stages of annealing. An obvious source of concern *with present models of annealing is the fact that, despite the conduct of the annual annealing procedure, the migration of fluorine and uranium from the fuel salt did occur. Once the reactive gas is removed from the off-gas system, simple experiments to heat the salt (even slightly) and observe evolved gas may shed additional light on gas diffusion mechanisms. Heating the snit with a controlled amount of fluorine (possibly as HF) in the cover gas might add a measurable amount of oxidant to the salt, aiding in the development of a more effective solid phase annealing procedure. Even if it is only limited in its effectiveness, implementation of a solid phase annealing procedure while taking the salt to near its melting temperature, followed by additional adjustment of redox potential in the liquid phase, may reduce the risks of unacceptable consequences during melting. 3.1.2.4 Adjusting the redox potential while melting Unlike adjustments to the salt redox potential in the solid phase, there is abundant experience with adjusting the potential in the liquid phase. Such adjustments were made routinely during preparation of the MSRE salts and throughout the salt operating phase. A common technique to add oxidant to the salt
3-8 is to sparge the molten salt with an HF/H 2 gas stream. To implement this approach, adequate assurance must be provided that undesirable consequences do not occur as the salt begins to melt before the redox potential has been adjusted. An approach to melting the salt while at the same time adjusting the redox potential is depicted in Fig. 3.4 (Toth and Williams 1996). The process uses a central heating system to melt a small controlled pool of salt at the top center of the salt mass. A HF/H 2 gas stream would be sparged into the melt to continuously adjust the chemistry through the reaction: This manner of adjustment of the fluoride salt redox state is based on proven chemistry. Any small amount of precipitate would settle to the bottom of the pool at the liquid/solid interface and would be subject to sparge gas contact without contacting anything but additional salt. The redox chemistry of the molten pool would be continuously monitored by analyzing the exit HF/H 2 ratio and comparing it with the entering gas stream. When the ratio in the exiting stream was equivalent to .that entering, the molten pool redox chemistry would be satisfactorily corrected and any reducing precipitate would be redissolved in the molten pool. Additional heat could then be supplied through the central heater, causing an increase in the molten pool. The previous pool salt volun1e would serve as a buffer to equilibrate with the newly melted salt and less severe inhomogeneities would result. Further sparging of the HF/H2 mixture would adjust the new pool volume to the desired redox state before the addition of more heat to the system. This process would be continued until all of the salt in the drain tank was molten. Some heat can be supplied to the overall system through the external heaters to raise the salt to a subliquidus temperature and thus lessen the heat requirements of the central heater. It is calculated that it would take
-5 days to adjust the fluoride deficiency in one drain tank with a 4% H2 in argon (non-explosive mixture) containing 4% HF flowing at l OL/min.
This approach to melting the salt would require additional heating and sparging equipment, but it can be integrated into the existing maintenance system that has been used to access the drain tanks in the past (Fig. 1.8). It is likely that heater elements and a sparge line can be designed to enter the central flange on top of the tanks. A new interface fixture between the existing maintenance shield and the drain tanks would be designed and procured. A potential heater system, based on the development of a rock melter at Los Alamos National Laboratory, has been identified (Jensen 1995). Other commercial heater systems may also be applicable. 3.1.2.5 Withdrawal of the molten salt Once the salt is safely melted, and the desired redox potential has been re-established, transferring the salt out of the tanks should not pose any major difficulties. Two basic alternatives exist. The first is to transfer the salt out from the drain tank cell into the fuel storage tank in the fuel processing cell, as has been done several times in the past. This requires pressurizing the drain tanks with helium. The integrity of seals provided by freeze valves and salt line plugs would have to be verified before this operation. Some indication of the condition of these components may be obtained when the off-gas system is evacuated by the reactive gas removal system. The condition of the salt lines themselves and the fuel storage tanks would also be assessed. The drain tanks are thick-walled, and not been exposed to corrosive gases at molten salt temperatures. Exposure to fluorine at the annealing temperature is not likely to have challenged the integrity of the drain tanks. However, the fuel storage tank was exposed to fluorine at molten salt temperatures when the uranium was removed from the fuel and flush salts in 1968.
HF/H 2 --- ROCK MELTER SPAR GE-- I i
,,./ MSRE DRAIN
_ ,,,,.-" TANK
,,/ MOLTEN ZONE V .~
I l-*m I ~"-~ * -J I
- SOLID * ~ l *
** ~....
- MSRE )----- =--__ * .\\ 1
- w
.I--- -
/
.______J:**
*ee
.... ***-" ~ * - * * - - I
- FUEL * \..._ ~./ * \0
**
- SALT * *****-..,.. .....-* _. .
*I
* .__ _ _ _ ___.
- EXTERNAL
- HEATERS - - - ~
I!
- 1) Heat supplied by external heaters to bring salt up to a subliquidus hot state.
- 2) Rock Melter supplies extra heat in specific region to cause localized melting.
- 3) Heat from Rock Melter increased as HF/H 2 sparge is continued.
Fig. 3.4. Diagram of process to melt a pool salt with an internal heater while restoring the fluorine balance with a HF/H sparge. 2
3-10 The salt could be transferred to the fuel processing cell for separation of uranium, or just as a step in removing the salt. A line extends from the fuel processing cell to the spare cell. In early facility drawings this line was indicated for the removal of salt; it was in fact used to transfer salt to the distillation e>..'J)eriment and remains connected to that experiment. The line could be cut and diverted to a salt canning station in the spare cell. Alternately, the molten salt could be lifted directly out of the drain tanks into a cask set up over the drain tank cell. This could be done by pressurizing the tank, or by lifting it with a vacuwn. Dependin g on the equipment arrangements, multiple vacuwn lifts could be required. Again, the general arrangem ent for this operation would be similar to that depicted in Fig 1.8. Withdrawal of molten salt from the tanks does not appear to present any fundamental challenges similar to the challenge of safely and effectively melting reduced salt in the first place. 3.1.3 Removal as Solid Salt As an alternative to removing the salt as a liquid, various approaches to removing solid salt were investigated. Three representative approaches were identified (Kring and Richards N.d.). These, involving a minor intrusion into the tank, a major intrusion into the tank, and removal of the tank and salt as a unit, are described in the following sections. 3.1.3.1 Removal with minor tank intrusion The removal of the salt through an opening in the top of the tank made by cutting out the 3 in. flange represents the least radical inln;1sion into the tank. It involves deploying hardware through the cell shield plug penetration into the opening. This approach would require removal of the steam dome and bayonets, and the thern1al insulation on top of the tanks. It also requires the adaptation or developm ent of presently available hardware and end effectors. The thimbles in the tank interior would be a serious obstacle to total mechanical removal of the salt. Any attempt to go through a small opening in the top of the tank is complicated by the fact that equipment must be small enough to deploy and then be navigated around the thimbles and other tank internal hardware. Options for removal included mechanical devices such as augers, impact chisels, chipping hammers, needle scalers or water blasters to break up the salt and then vacuum or scoop the material out of the tank. A CO blaster that would 2 dislodge or break up the salt with simpler mechanical hardware was eventually selected as the desired option to pursue. The minor tank intrusion sequence is illustrated in Figs. 3.5 and 3.6. These figures identify the equipment and facility modifications necessary for the operation. The operation of the CO/vacu wn ablation wand will require the installation of a remote salt collection system consisting of a recircula ting loop of air provided by a vacuun1 suction nozzle, cyclone separator, bag filter, vacuwn blower, and return air piping loop to the tank interior through a primary containment extension. Collected salt will be dispensed into appropriate containers and stored in the drain tank cell. Two approaches are shown for deployment of the CO/vacuwn wand. A manipulator-actuated approach is shown in Fig. 3.5, and a stand-alone assembly is depicted in Fig. 3.6. The manipula tor-actuated wand nozzle is positioned to direct a stream of high velocity solid CO pellets the size 2 of a pencil lead to impinge onto the salt. The salt is abraded off into particles to be swept into the vacuum annulus of the nozzle and transferred to the cyclone separator. The salt particles are collected in containers located in the drain tank cell. The wand (Fig. 3.7) is attached to a tank primary containm ent e>..iension that prevents distribution of salt dust throughout the drain tank cell. The wand is spherical ly
THINBL[ CO2 WAND NT IN SPHERICAL BRNG 02/VACUUN NOZll[ 11TH 2 O[G or rRCCDON w UCL DRAIN JANK I IIJH SON[ JHINBLES SHOIN
*, FUEL SALT EICAVAT[
*, DOIN AND our TO 1ST
\ THIMBLE ROI i
I
/
20 KG SALT STORAGE vcsm
.~*<'\_SE C DETAIL A DETAIL A Fig. 3.5. Depiction of the hard war e need ed to mechanically remove the salt usin g a CO2 blas ter mounted to an overhead man ipul ator system.
(" .. ,-
3-12 VERTICAL NAST ACTUATOR PRIMARY CONTAINMENT EXTENSION NUL Tl
- DOF AR DEPLOYED CO2 ABLATOR TANK HANGER RODS 3 IPS TANK EXTENSION REMOVED AND HOLE ENLARGED TO 8 IPS 02/VACUUN NOZZLE 11TH 2DOFS SUPPORT POLES EATER JACKET UEL SALT DRAIN TANK WITH COOLING SYSTEM TOP HEATER JACKET AND IEIGH CELLS REMOVED Fig. 3.6. Depiction of the hardware needed to mechanically remove the salt using a CO2 blaster mounted directly onto the drain tank.
~2 1AM0/M AMIPULA10 R AOAl'lER \\ SPHERICAl CO2 PEllE BHEARtllG OUSIMG
\ AMO COM l PRESSi.0 AIR (.ONM i.tT\Oll RIMAR1 CO UlEKSIOlt MTA\MMEMT \ '\ 1\ ,\ ,\ 0 *VACUUM tO 80 IM 1A l0 M\tECl \011 v.> v.> SECllOM IH \ \ 2 OOf MO llll-Fig, 3,7, De~iction o! t b
- wan d and no,:
r.le for tbe C O , abla tion ~,oce ss,
3-14 mounted, providing both vertical roll and pitch, allowing the nozzle to target the space between the thimbles. Viewing equipment will be inserted into the tank to observe the process. The stand-alone option (Fig. 3.6) is more complex but has no dependence on the pendulum characteristics of the telescopic mast manipulator. In both cases, an initial excavation of the salt down and out of the thimble radius may be required to provide access for wand installation and to decrease the overall time required for salt removal. The rate ofremoval of the fuel salt using the CO2 ablation technique is not known. A development program would be required to prove the technique in general and to demonstrate the needed dexterity to position the nozzle between and around the thimbles. The installation will require removal of the existing 3-in. nozzle on top of the drain tank and replacing it with an 8-in.-diam. tank containment extension for mounting the wand. As a result, remote cutting and possibly remote welding will be needed to prepare the equipment. This, in turn, may require a special cell membrane extension and a glovebox-type secondary containment to operate and install the cutting and welding systems. 3.1.3.2 Removal with major tank intrusion Removal of the salt after cutting off the top of the tank represents the most radical intrusion into the tank This approach involves the greatest number of remote activities of all the mechanical removal options. Its single advantage is the exposure of the salt bed, free of obstructions for salt mining operations. This approach involves deploying hardware through the access shield plug penetration into the opening. It requires removal of the steam dome and bayonet assemblies, as well as the thermal insulation on top of the tanks. An extension of the tank shielding would also have to be built. Remote cutting tools, based either on core drills or a plasma torch, would be used to remove the thimbles and/or the upper dished head. Options for obtaining access to the tank would be to individually separate and extract the thimbles or to separate the thimbles from the salt while still attached to the tank head. The extraction of thimbles from the salt is complicated by the bond between the salt and the thimble surface. If this bond proves to be weak, rotary or impact shearing techniques may prove adequate to free the thimble. Applying heat to the thimble may serve to weaken the bond. At worst, the thimble could be drilled out with the salt. After the top of the tank is removed various tools could be used to remove the salt Adaptation of commercially available tooling would be preferred over the development of new types of tooling. These operation are depicted in Figs. 3.8 and 3.9. Figure 3.8 shows the salt being removed using a dedicated snit milling/vacuum device mounted on the tank, and positioned by a general-purpose polar fixture. The system consists of a radial arm, rotary mast milling head and vacuum auger that provides for the vertical, radial, and rotary motions of the auger, and a salt collection system. The latter includes a nozzle, a cyclone separator, a bag filter, a vacuwn blower, and a retum line to the tank. Viewing, handling and maintenance of the equipment is provided with a remote manipulator mounted on a telescoping mast crane located in an extension of the secondary containment above the cell. After most of the salt has been removed, a final cleanup system will be needed to remove salt crust from the tank interior. Equipment for this purpose could include a CO2 blaster, a needle descaler, or a rotary descaler. Vacuum transfer of the salt would be accomplished separately or using a vacuum system integrated into the tools. A development program would be required for all of the tooling described in this section.
i I' cp( JI_ CUUM VENfURI YCLONE SEPARAJOR DETAIL II SCALE o.m ACUUM INTAIE fOR GRANULATED SALT ILLING HEAD w I I Vi fUEL DRAIN TANK 1 i I I SEE DETAIL '\ SALT CONTAINER Fig. 3.8. Depiction of a system to mechanically remove the salt from the tanks using a vacuum auger.
~
~
,,,,,. .. - .. - .. _ / S E E DETAIL A
/ j
/ ~ *, 9
( ~ L_ J L_R___R__I_ J \)
/
~ w
/
/ Rill &GUIDE - - - -
I
°'
/
**, ,,/
I i ' c::;:1.f PPER
'HEADI DI SHED
- 1,
*SPACER R'I NG
~ DETAIL 8 OWEA DISHED HEAD DETAIL A I
""'-- SEE DETAIL B SECT ION A*A Fig. 3.9. Depiction of a system to with draw the cooling thimbles from a fuel salt drai n tank .
3-17 3.1.3.3 Removal of salt and tank as a unit The tanks, with the solid salt in them, could be removed as a unit and transferred to some other processing facility for removal of the salt. or they could be lifted into an extension of the drain tank cell to enable mechanical access to the tank interior with shorter tooling. In either case, a basic set of activities is required. In the case of a transfer, the tank is lifted into a cask (Figs. 3.10 and 3.1 J). In the case where the tank is lifted to perfonn additional mechanical operations, the tank would be lifted into a cell extension provided with shielding windows and manipulator systems. To remove the tanks, all instrumentation and heater lines would have to be disconnected or cut. The steam dome and bayonet assemblies would be freed from the water and condensate lines and the tank supports, and would be withdrawn through the maintenance shield into plastic bags or similar enclosure s to prevent the spread of contamination. Salt and off-gas lines to the tanks would be cut and plugged. A special, bottom-loaded shielded cask would be procured. A structure would be erected on top of the drain tank cell (presumably at the high bay elevation) on which the cask would be set. This structure would complete the shielding around the cask. A contamination control barrier would be ~reeled around the path that the tank travels as it is lifted into the cask. A lifting fixture would be attached to the tank, and the tank would be lifted using a hoist that is integrated into the top of the cask. Once in the cask, the tank would be placed on a jig to provide support for the tank. The cask bottom shield would then be closed and the cask can be moved from the drain Lank cell. Provisions would have to be included in the design of the cask system to control reactive gases from the irradiated salt in the tank. Once in the cask, the tank could be moved within the ORNL site, or it could be loaded into a cask approved for transport over the road. At present. no existing, licensed cask has been identified. A loading pit, would have to be identified for the transfer, and tank supports would have to be integrated into the transportation cask. It is unlikely that the cask depicted in Fig.3.11 could be licensed for transportation over public roads~ even if so, the cost of obtaining such a license would be high. 3.1.4 Removal with a Liquid Solvent/Carrier 3.1.4.1 Dissolution The most obvious, common solvent for the MSRE fuel and flush salts is water, or water with various acidic or basic additives. Because uranium generally precipitates out of basic solutions, acids were the focus of this evaluation. A set of simple dissolution experiments were carried out to establish the solubility of the MSRE fuel and flush salts (personal communication, L. M. Toth, September 6, 1995). The results of these tests are shown in Table 3. I. Table 3. I. Solubility of components in MSRE fuel salt in various solvents (g/L at 25°C) Solvent medium LiF BeF2 ZrF4 HP 5.6 2.7 25 4MHF 2.9 5.0 JO 4MHF+ 2MHNO3 3.3 2.5 4.2
3-18 r---------SPEC IAL SHIELDED CASK CASK SUPPORT PLATFOR AND CONTAMINATION BARRIER FUEL SALT DRAIN TANK WITH LIFTING FIXTURE AND JIG NSRE FUft SALT DRAIN TANK CE Fig. 3.10. Depiction of the general arrangement of equipment and shielding needed to remove a fuel drain tank and salt as a unit and transport it to another location.
I
- i I
I I TANK LIFTING FIXJURE XISTING TANK ASSEMBLY JIG SPECIAL TANK SHIPPING wI CASK
\0 ASK SN IELD DOOR FUEL SALJ DRAIN TANK Fig. 3.11. Depiction of a cask system that could be used to remove a fuel drain tank and salt as a unit and transpo rt it to another location.
3-20 At these low solubilities, it would require more than 500 m3 of water to dissolve the salt, resulting in a volwne increase of more than 250 times. The salt did not dissolve readily. The salt was broken up to increase the surface contact with the solvent, and was agitated. These experiments indicated some of the difficulties that would be encountered with salt dissolution, based on typical aqueous solvents. Further complications result from the need to maintain criticality safety. Simple models show that if the salts have a void fraction of 2%, and that void is filled with water, a nuclear criticality is likely. Thus, any solvent may need to incorporate a neutron poison. It might be necessruy to drill a core through the salt, add a measured amount of water, and withdraw the water from the bottom of the tank before more can be added. If the water is taken up in the salt matrix (either incorporated into the salt itself or trapped in cracks and fissures in the salt monolith), this procedure may not work. A need to draw water from the bottom of the salt, using a core drill, incorporates some of the concepts of mechanical removal into this approach. 3.1.4.2 Suspension in a slurry An approach to transport the salt with a smaller amount of solvent would be to break up the salt and transport it as a sluny. In a sense, this is the same concept as breaking up the salt with a CO blaster 2 and vacuuming up the broken particles, except that a liquid medium is used instead of the first solid, then gaseous, CO2 medium. Thus, this approach has in essence been covered in Sect. 3.1.3.1, and is not addressed further here. 3.1.5 Screening of Removal Technologies The technologies identified for removal of the salt include approaches for removal as molten salt, mechanical removal of the salts from the tanks, removal of the salt as an aqueous solution or sluny, and removal of the salt and tanks as a unit. Of these, removal as molten salt is the most straightforward and requires the least additional equipment. It is likely to be far less expensive than other options. But chemistry issues associated with excess reductant now present in the solid salt as a result of the migration ofF2 and UF6 from the tanks must be resolved before the tanks are melted. Techniques for melting the salt while safely adjusting the chemistry have been proposed, and an experimental program to verify these techniques has begun. Removal as molten salt, with properly adjusted redox potential, interfaces well with useful technologies for separation of uraniwn and other materials from molten salt. Removal of the salt by mechanical means or as a solution or sluny is more expensive, entails the development of more and complex equipment, and is likely to incur more cost and risk of the spread of contamination. Removal using the CO2 blaster appears to be the most practical of these, and could be a viable backup to removal as a molten salt. Aqueous removal techniques are hindered by the very low solubility of the salt, and by the potential for a nuclear criticality if water moderator is introduced directly into the salt matrix. Removal of the tanks filled with solid salt is possible, although the cask and tooling needed for this operation would be expensive. To date, a destination that could accept and further process the salt and tank as a unit has not been identified.
3-21 3.2 SEPARATIONTECHNOLOGIES 3.2.1 Fluoride Volatility Fluoride volatility is a technology developed to separate uranium and other materials from fluoride salt solutions. As such, it has been applied to molten salt reactor fuels throughout the development of such reactors. Uranium was separated from carrier salt used in both the MSRE and its predecessor, the Aircraft Reactor Experiment. The operating experience with the MSRE salts is well documented (Lindauer 1969), and an overall review of the technology has recently been completed (Advanced Integrated Management Services I 996). The flowsheet used to separate uranium from the MSRE is shown in Fig. 3.12. Salt was transferred from the drain tanks to the fuel storage tank in the adjacent fuel processing cell. This tank resembles one of the drain tanks in many ways, including the overall configuration of being hung in a furnace on a support ring that in turn rests on weigh cells. The tank is slightly larger than the drain tanks, so that the entire inventol)' offuel or flush salt can be fluorinated in one batch, with sufficient freeboard in the tank to minimize salt caeyover in the outlet gas. The outlet gas then passes through a high-temperature [399°C (750°F)] NaF trap located near the floor of the fuel processing cell. Volatile fission products were collected on this trap, but at this temperature uranium passed through. Uranium was then collected on a set of NaF traps located in a cubicle in the high bay. Excess fluorine was then reacted with KOH in a liquid caustic scrubber, and the outlet from the scrubber passed through a mist filter, a soda lime trap, several stages of charcoal, and a HEPA filter before being discharged to the MSRE ventilation system. The salt was transferred at a relatively high temperature, and was then cooled to within 10°C (50 °F) of the liquidus temperature to minimize corrosion and fission product volatilization during fluorination. The sample line was purged with helium to prevent condensation ofUF6 at the cold upper end. The salt was sparged with either pure fluorine or a fluorine-helium mixture at a relatively high flow rate (~40 L/min) to convert the UF4 to UF5* Fluorine utilization was high during this period. When all the UF4 had been converted to UF5, UF6 began to form and volatilize, as indicated by a temperature rise in the first absorber and a rise on the inlet flowmeter. When this *occurred, the fluorine flow was reduced to 15 or 25 L/min to increase the absorber residence time for more efficient absorption and to increase the fluorine utilization. The gas leaving the fuel storage tank consisted ofUF6 , excess fluorine, helium, MoF6 and some CrF4 or CrF5 from corrosion, IF,, and the fluorides of some other fission products such as tellurium, niobiwn, ruthenium, and antimony. The gas passed through the 399°C (750°F) trap where the chromium and most fission products were absorbed. A small fluorine flow was introduced upstream of this bed to ensure an excess of fluorine and to prevent any nonvolatile UF6 from forming and remaining on the heated NaF in the event that the fluorine utilization was near I00%. Although the fluorine utilization was not high enough to require this additional fluorine stream, a small flow was maintained to prevent diffusion and condensation of UF6 in the line. The gas stream then left the shielded cell and passed through five NaF absorbers in a sealed cubicle in the operating area. These absorbers were heated to 93-121 °C (200-250°F) to increase the reaction rate and to minimize MoF6 adsorption. The UF6 piping was heated to at least 60 °C ( 140 °F) to prevent condensation of UF6
- As the UF6 began to load on a given absorber and the temperature started to rise, the cooling air was turned on that particular absorber to limit the temperature to a maximum of l 77°C (350°F). High temperature reduced the uranium loading by promoting surface absorption and reducing penetration of the UF6 to the inside of the pellets. The final absorber was operated below 121 °C
ORNL-OWG 68-8994 sAtA~{ER CH~~~TNG I Hl2~EiAY ~ i===i 2OO°F NaF ABSORBERS IN CUBICLE
=== == ~
He AND FUEL PROCESSING 7in
=tr ACTIVATED CHARCOAL ------c.>.:: ::--:-i"F. CELL U TO SALT TO OR ~LIQUID TRAP WASTEtA FROM DRAIN l I ~
AND FLUSH FU,EL ff--!--1 I MIST TANK fs TO VENT SYSTEM TANKS STORAGE ! ~ FILTER ~ vJ I Iv Iv SODA
~ LIME TRAP WASTE 75O°F CAUSTIC CHARCOAL TRAP SALT NaF BED NEUTRALIZER TANK Fig. 3.12. Flowsheet of the MSRE salt processing system.
3-23 (250°F), where the partial pressure ofUF6 over the UF6 *NaF complex allowed only a negligible amount of uranium to reach the caustic scrubber. If the gas stream leaving the absorbers contained greater than 50% fluorine, it was diluted with helium before reaching the caustic scrubber. This allowed a smooth reaction in the scrubber without the production of a flame or pressure oscillations. The caustic scrubber was charged with 1300 L of 2 M KOH- 0.33 MKI containing 0.2 M K2B~O1, which was added as a soluble neutron poison. The reaction that occurred in the scrubber is The scrubber solution was replaced before one-half of the KOH had been consumed, as detennined by fluorine flow and calculated utilization, because dip tube corrosion was increased when the OH-concentration was less than l M. Besides fluorine, most of the molybdenun1 and iodine were removed in the scrubber. The high-surface-area mist filter localed downstream of the scrubber removed any particulate matter from the scrubber. The soda-lime trap provided a means for detecting fluorine and a means for removing traces of fluorine from the scmbber off-gas before it reached the charcoal absorbers. Activated impregnated charcoal traps sorbed any iodine that was not removed in the caustic scrubber. The gas leaving the charcoal traps consisted of only heliun1 and o:-..-ygen that was produced in the caustic scrubber. This gas flowed through a flame arrester and then through an absolute filter before being routed to the cell exhaust. The flush salt was fluorinated and the uranium was trapped in a single set of NaF absorbers. After 5 h of fluorine sparging the flush salt contained 24 ppm uranium. The salt was sparged for an additional I hand 49 min at which time the uranium concentration was 7 ppm. A total of 6.5 kg of uranium was recovered from the flush salt. The average fluorine utilization was 7. 7% during fluorination of the flush salt. Fuel salt was fluorinated in a single batch, and the uranium was trapped in six sets of fluorine absorbers. Between runs, the fluorine trailers were changed and the caustic scrubber was recharged. Approximately 7.5 h of fluorination at about 40 Umin was required for conversion of UF4 to UFs before UF6 evolution was observed. After UF6 evolution began, the fluorine utilization remained steady at about 30% for - 33 h of fluorine sparging at a rate of about 19 L/min for an average fluorination rate of about 6 to 7 kg/h. During the last 6 h of fluorination, the fluorine utilization decreased from about 30% to near zero as the uranium concentration in the salt decreased to 26 ppm. A total of -45 h of fluorination time was required to fluorinate 216 kg of uranium from the salt. The overall fluorine utilization was 39% during fluorination of the fuel salt. Identifiable uranium losses were less than 0.1 %. The recovered uranium was decontaminated from fission products by gross gamma and gross beta decontamination factors of8.6 x 108 and 1.2 x 109, respectively. Use of this existing equipment is being evaluated for separation of the 233U now present in the fuel and flush salts (DeVore N.d.). Certain changes would be required to accommodate highly enriched material (the original charge was about 30% 235U). The off-gas system would need to be evaluated in tenns of providing effective retention of 232U decay products, including radon. The fission product inventory is now much less significant, and use of the hot NaF trap ahead of the uranium absorbers may not be necessary. Because of the radiation associated with :mu daughters, including 208Tl, shielded (and critically safe) absorbers would be required. Instrumentation would have to be upgraded, and the physical integrity of all components would have to be verified. This includes the integrity of equipment in the
3-24 drain tank cell, to ensure that the drain tanks can be safely pressurized to transfer salt to the fuel processing cell. If the existing salt processing cannot be reused (at least, in its entirety), several other options exist. The salt could be immediately withdrawn into the containers that have been described earlier and moved into a shielded location for fluorination. The gas outlet from the container would pass through the new shielded absorbers and into the off-gas system. The existing caustic scrubber and the rest of the existing off-gas system might be used in conjunction with the new trapping station and fluorination system. This alternative would be well-suited to a salt removal approach in which molten salt is withdrawn directly from the drain tanks. It avoids the need to certify all of the drain tank off-gas and salt piping systems for pressurization of the salt, and avoids use of the fuel storage tank. This tank was subjected to significant corrosion during the 1968 fluorination operation. The fluorination process was also evaluated for separation of plutonium from the salt (Advanced Integrated Management Services 1996). Results from laboratory and pilot plant studies indicate that plutonium is more difficult to fluorinate than uranium and very little volatilization of plutonium occurs until most of the uranium is removed. Long fluorination times may be required to significantly reduce the plutonium concentration. This also increases the total amount of corrosion experienced by the fluorination vessel. Plutoniun1 is also thenually unstable at elevated temperatures and a high fluorine to PuF6 ratio is necessruy to prevent decomposition. If the plutonium is removed by Ouorination, it is likely that some decomposition will occur that will result in deposition of PuF in process piping and vessels. 4 Thus, the fluorination process is not recommended for separation of plutonium from the salts. 3.2.2 Electrometallurgical Separation The electrochemical process for treatment of MSRE fuel and flush salts is being developed and demonstrated (with sinmlated MSRE fuel salt to which representative fission products have been added) by Argonne National Laboratory (Laidler l 996a,b). The purpose of the electrochemical treatment is the extraction of actinides and fission products from the fuel/flush salt, concentrating the radioactive constituents in minimal-volume high-level waste fonus and leaving a large volume fluoride salt residue containing only trace amounts of radioactivity. The process is carried out in a very compact system, with the present concept for fuel/flush salt treatment involving 30-L batches of salt (a slightly smaller quantity than the salt container described in earlier sections, which divides each of the fuel and flush salts into 24 containers *with about 80 L of salt in each). Electrochemical treatment is conducted in an inert atmosphere glovebox, about 25 ft long by 8 ft wide by 14 ft high (Fig. 3.13), which can be installed in an existing hot cell or be enclosed with temporary shielding. With the envisioned 30-L batch size, the shielding requirements are about 6 in. of lead.
- The process is described in a series of steps, as shown in Fig. 3.14. The materials that result from this process, along with a possible disposition, is indicated on Table 3.2. The quantities indicated are for the entire batch of fuel sail, roughly 2000 L. The process is carried out in 30 L batches, or about 65 batches. Each batch would be in process for one week. The processing approach is to displace the radioactive cations in the MSRE salt with nonradioactive alkali metal (either Li or Na) cations.
In step A of the process, zirconium would be removed from the salt and replaced with lithium. Product 1 of the process is zirconium metal dendrites, plus metals that are more noble than zirconium. The zirconium is a component of the original fuel salt; the other metals would come from any noble metal fluorides that were present in the fced salt and the metal daughters from the decay of 228 Th that are more noble than zirconium. This stream is initially strongly radioactive, but since most of the activity
2.4m (8 rt)
/f P,5m
( 14_ /"I) w I N VI I
- j I
I Fig. 3.13. Overall dimensions for a shielded inert atmosphere enclosure to house the Argonne electrorefiner. r
3-26 ANODE ~ ~ - - - - - - - . 136 kg Li m--r ----- ---+ ----* ttt-m -- NOB 1-FP METAL WASTE (70L ) LE METAL FPS AND EQUIV. Th DAUGHTERS; 5: 7 kg; Zr-4 45 kg
-A SOLID CATHODE SALT SHIPPING CONTAINER PROCESS CRUCIBLE HEATER ASSEMBLY MSRE FUEL SALT 4650 kg: 1996 L Li/B e/Zr : 65/ 31/ 4 ANODE 4 kg Li EQUIV. Bi POUNDER Bi 3-TW O .30L BATCHES GATHODE OF CL ELECTROLYE WITH 0.36 kg TRU /RE- TR ANODE -++r.1.:-.-. - - - - . , . , .
<1 kg L; ut-" --t-- ----- ;.,= .nr- - BIS MU TH
~:JOO Kg (104 L)
EOUN. Re,C s-Ba Th-DAUGHTERS TRU -0.3 6 kg: 61.2 kg FPS ACTIVITY - 60,SOOCi H35w POUNDER Bi CATHODE ALLOY .w.=-,=------. 5-ME TAL WASTE (100 L) ANODE it-+ ---- ---+ :::: m- --S r/Y ACTIVITY 15,0 00C i 50w D POUNDER LIQUID METAL CATHODE 6-SA LT WASTE - LiF-BeF2 430 0 kg, 2075 L, 23m Ci, I Fig. 3.14. Diagr am of the processing steps used to electr ometallurgically separ ate radioactive components from the fuel salt.
Table 3.2. Process steps and products for electrochemical treatm ent ofMSR E fuel salt Step Product Volume 0 Composition I Mass0 Disposition
- I A Fission product metal waste 70 liters l - NM fission products and Th daughters ..
(Mo, Ru, Pd, Cd, Rh, Tc, Se, Sb, Nb, 6kg HLW; combine with metal waste Zr, Po, Bi, Pb, Tl) from treatment of EBR-11 spent fuel
- Zirconium metal from fuel salt .............. 455 kg Bb (In-Process) Bismuth ingot n/a - Bismuth ingot containing: In-process material; sent to Step B-1
- uranium .............................................. 37 kg
- transuranic elements .......................... 0.4 kg
- rare earlh fission products .................. (trace)
B-lb High-purity uranium ingot 2 liters 233 U and trace amounts of 232 U .............
- 37 kg Return to ORNL 233 U repository Chloride salt waste 1.4 liters - Transuranic elements ........................... 0.4 kg HL W; incorporate in EBR-11 w ceramic waste form t!..>
-..:i
- Rare earth fission products, Cs-Ba, Th, Interim storage for decay of Th; 228 C Bismuth ingot I 04 liters Th daughters ......................................... 6kg final conversion to I IL W disposal
- Transuranic elements ........................... 0.4 kg form as glass-ceramic
- Bismuth (non-radioactive) ................... 1000 kg Metal waste ingot l00 liters - SrN fission products ............................ <4 kg HLW; combine with metal waste
- Metal ingot ........................................... 600 kg from treatment of EBR-11 spent fuel I D I Fluoride salt residue 2075 liters - LiF/BeF2 ................................................................................ 4300 kg LL W; possible conversion to nuorapatite before disposal
- Based on treatment of2,000 liters ofMSRE fuel salt b If uranium is initially removed by fluoride volatility process, these steps are omitted
-1 I I
3-28 is from ~I it *will decay away in a week. This metal waste could be made part of the metal waste fonn generated by the electrochemical treatment ofEBR-II fuel at ANL-W. In step B, uranium is removed from the salt by electrochemical titration with lithium into a liquid bismuth cathode. the use of bismuth in the cathode provides the means for removal of the active metals from the salt, yet leaving the active metal Be in the salt as a salt constituent BeF
- While removing all 2
of the uranium into the cathode we would remove about half of the plutoniwn and a trace of the rare earths. To make a high-purity uranium from this bismuth metal cathode, it is made into the anode of a second cell (which uses LiCI-KCI as the electrolyte). High purity uranium metal (product 2) is produced at the cathode of this second cell. This metal could be sent directly to the ORNL RDF for storage, or could be converted to oxide for storage. There is no physical connection between the fluoride and chloride cells. Plutonium and the rare earths accumulate in the chloride cell electrolyte, as they do in processing the EBR-11 fuel at ANL-W. The salt left over at the end of the campaign is thus a candidate for further treatment at the EBR-Il facility. The total quantity of chloride salt is at most 60 L, or less than half the salt charge at the ANL-W facility. At the completion of the MSRE campaign the process crucible from the chloride cell would be sealed and shipped to ANL-W, fanning product 3 of the process. If uranium were separated by fluoride volatility before placing the salt in the fluoride electrorefiner, the second chloride electrorefiner would not be required. There would only be one extraction into bismuth, represented by the following step. In step C, the remainder of the plutonium, rare earth fission products, the 137 Cs/137mBa, and the thorium (including daughters of 228Th decay) are removed from the salt into a bismuth cathode. This fomJS product 4 of the process. It is reconunended that this metal product be held in interim storage until the 228Th decays, with a 1.9 year half-life. A number of options are available for the ultimate disposal of this waste: (I) qualify the bismuth metal waste directly as a waste fonn. (2) since Bi is "glass forming," the entire waste could be added to a glass melter, or (3) the bismuth could be electrorefined to separate the rare earths and cesium from the bismuth, leaving a chloride waste salt that might be disposed of at ANL-W. In step D, the strontium (including 90Sr) would be electrorefined out of the salt. This will require the development of nitride- or oxide-enhanced technique for the extraction of the strontium at the cathode, since this ordinarily does not happen. The strontium waste fonn,produc t 5, is a small bed of nitride or oxide solids trapped within the cathode metal ingot. Candidate metals for this cathode are bismuth, cadmiwn, lead, antimony, or alloys of these metals. Since 90Sr has a half-life of 28 years it could be placed into storage for 150 to 200 years for decay. Finally, product 6 is the basic matrix salt LiF-BeF2 (or NaF-LiF-Be~) stripped of actinides, transuranics, and fission products. The salt volume is essentially unchanged from that initially fed into the process. This residue might be disposed of as low-level radioactive waste. The mechanical handling steps are depicted in Figs. 3.15 through 3.17. Figure 3.15 shows the crucible handling steps, including bringing the salt can in through the transfer lock, operating the electrorefiner with the crucible in place, and removing the crucible after processing of that batch of salt is completed. The second step depicts handling of the solid cathode used for zirconium removal. Activities shown include steps used to separate salt from the metal product, including processing with a mandrel stripper and a spinning operation. The product receiver is shown leaving through the transfer
\I
.I REMOTE OPERABLE ROBOTIC GANTRY 9 AXIS W/INCHANCEABLE HANDS CRUCIBLE WITH LID READY roR TRANSFER SALT CAN R£AOY t,J FOR INSERTION INTO REACTION VESSEL N a":i;1.:*~£t::~* r:fil,, \0 ill
~*.:-..!:~~:.C,~!;
- if{ri~~~~{
ltfil ... &!le
*****l II
~::::::=..__ .l_____
I I Fig. 3.15. Depiction of the crucible handling activities for the Argonne electrorefiner.
CATHODE UANDREL smlPPER
':\
ROTATABLE COVER RECIE.\IER U ALlOY ANODE AND REPLACE IN SHIPPING CASK l,J w 0
~ 1111 II ' II 7/ " _;:-.:::.*.*.--
ANODE CATCH BUCKET REACTION VESSEL soo* - 600* c Fig. 3.16. Depiction of the solid cathode handling activities for the Argonne electrorefiner.
\
- 'j Ii I BISIJUTH INGOT PREPARATION SHIPPING CONTAINER TILT POUR READY FOR fURNACE LOCK-OUT wI BISMUTH INGOT w
I~ I II I / '* II .... ****** ..... II
--****-*. ~~~~--r1,___, REACTION VESSEL 500° - 600° C Fig. 3.17. Depiction of the bismuth electrode preparation and salt cleaning from the pounder cathode steps activities for the Argonne electrorefiner.
3-32 lock Figure 3. I 7 depicts the handling steps for the liquid metal pounder cathode. An ingot preparation furnace is shown. During operation, the pounder mechanism agitates the metal-salt interface to "pound" the cathode deposit into the metal, ensuring that it goes into the metal solution instead of depositing on the cathode surface. After operation, the metal cathode is cooled and the solidified bismuth ingot is removed. The electrorefining process is well-matched to a starting point of a halide salt containing uranium, actinides and transuranics, and fission products (in the EBR-11 process, mechanical segmenting and dissolution activities are needed to get the. fuel into the chloride salt). It produces a good-quality uranium product and, with the addition of the nitride- or oxide-enhanced strontium removal step, decontaminates the salt matrix sufficiently to dispose of it as low-level waste. However, the other product streams are not readily disposed of at ORNL. These include the zirconium metal waste, a chloride salt contaminated with plutonium and other radioactive materials, a waste fonn that incorporates plutonium and other materials in a bismuth ingot, and a metal waste incorporating the 90Sr. If the uranium is removed by fluoride volatility before electrorefining, the second chloride electrorefiner is not needed. If the uranium could be disposed of with the other materials in the bismuth, the chloride electrorefiner is again eliminated. Storage of the salt with 90Sr until the strontium decays could be considered, eventually resulting in a proposed low-level waste form. This would reduce the waste products to the zirconium metal waste (a material that must be remo,*ed before other materials are removed from the salt) and the wastes encapsulated in bismuth. Ensuring the disposal of the bismuth product is a key to evaluating this technology. If the waste were qualified for direct disposal, this would make electrorefining a more attractive technology. Likewise, if a bismuth glass were qualified, the process would be attractive. If the full process were to be implemented, it would be useful to implement it at ANL-W, and directly interface with the EBR-11 electrorefiner. The full strategy described above relies on several types of material being transferred to Idaho in any case. Issues with waste qualification were identified in a recent National Academy of Sciences review of the electrorefiner process (Committee 1995). Costs of constructing an electrorefiner at ORNL are not thought to be very high (a few tens of millions), but the costs of the inert atmosphere facility are not included. At present, an appropriate facility at ORNL has riot been identified. 3.2.3 Vacuum Distillation of Salt Low-pressure distillation was studied as a method for separation and recovery of LiF and BeF2 from less volatile rare earth fission products in the Molten Salt Breeder Reactor (MSBR) program. Removal of the rare earths was desirable in order to achieve a high breeding ratio in the MSBR. It was not considered economical to discard the salt to remove rare earth fission products in the MSBR. Investigations into low-pressure distillation included the operation of a test still using both nonradioactive salt and fluorinated MSRE fuel salt (Hightower et al. N.d.; Advanced Integrated Management Services 1996). This experiment, located in the spare cell adjacent to the fuel processing cell, is shown in Fig. 3.18. Low pressure molecular distillation used for volatilization of the molten salt is considerably different from conventional distillation systems in which bubbles of vapor occur in the bulk of the liquid. In molecular distillation, vaporization occurs from the quiescent liquid surface and the vaporizing component is transferred to the surface by diffusion or convective mixing. Evaporation rate is proportional to the surface area exposed. Molecular distillation requires large vapor handling areas with the condenser located within a few mean free path lengths of the evaporating surface. These requirements
0RNL DWG 70-6280 ARGON SUPPLY PRESSURE MEASUREMENT9-s-AND CONTROL ., HCV-2 I I VACUUM FILTER PUMP
,------- ------, I ARGON I FILTER SUPPLY~ I LEVEL .j MEASUREMENT 1 AND CONTROL I
FREEZE I
-1 VALVE I I
I I SAMPLER H2 I w I w w MOLTEN SALT FILTER FUEL STORAGE TANK FEED TANK CONDENSATE RECEIVER Fig. 3.18. Simplified flow diagram of the MSRE distillation experiment.
3-34 make scale up or the design of multi-stage units difficult. Data from operation of the single-stage unit using MSRE type salt indicated that distillation rates on the order of 1.5 ft3 of salt per day per square foot of vaporization surface can be achieved. Operating experience with the MSRE still (0.52 ft2 evaporating surface area) included 6 test runs with nonradioactive MSRE-type salt for a total operating time of -293 h during which time about l 85 L of salt was distilled. In the test with radioactive salt from the MSRE (in the same still) - l 2 L of salt was distilled in 23 h of operation. Operating difficulties during the period with nonradioactive salt included plugged off-gas lines due to ZrF deposits and 4 plugged feed lines due to metal deposits. The vaporization rate may have been reduced by concentration polarization and droplet entrainment in the vapor may have occurred. The corrosion rate of the vessel appeared to have been acceptably low. Single state low-pressure distillation of the MSRE fuel and flush salt will likely produce an overhead product containing most of the salt and volatile fission products (mostly cesium fluoride) and a bottom product containing salt with a higher LiF content that the original salt {possibly 3% to 5% of the original salt volun1e). that contains most of the uranium, plutonium, and major fission products other than cesium. Based on the work done for the MSRE off-gas uranium, the uranium from the distillation will require considerably more processing before it is suitable for long-term storage. It is not clear that the storage capability of the overhead product {possibly containing 95% of the salt) has been significantly improved (other than from removal of the uranium and plutonium) since it still contains cesium, is highly radioactive, and contains a high concentration of fluorides. Removal of enough of the uranium, plutonium and americium from the overhead product could lead to the larger quantity of salt being classified other than transuranic waste. While distillation rates and corrosion data from the short tem1 engineering scale distillation tests indicate that distillation of the quantities of fuel and flush salt to be processed is probably feasible, it is expected that the processing rate would be slow and considerable design and development work would still be required. In summary, low pressure distillation might provide a technology that concentrates uranium, plutonium, strontiun1 and other radioactive materials in a small quantity of salt, while leaving the cesium and other radioactive materials in the larger quantity of salt. It is not clear that this technology, by itself, accomplishes significant gains toward disposing of the fuel or flush salts. 3.2.4 Other Uranium Separation Technologies A number of other alternatives for separating uraniUill from the salt have been identified and briefly evaluated (Advanced Integrated Management Services 1996). None of these was found to offer any advantages over the technologies already described. 3.2.4.1 Precipitation of uranium from molten salt Conversion of the uraniU111 into a compound that is insoluble in the fuel salt could allow separation of the uranium as a solid. Various strategies to precipitate uranium were identified. No insoluble fluoride was identified. Several strategies were identified that might convert uranium in the salt to an oxide. However, removal of the insoluble oxides from the salt poses several practical difficulties. This technique was considered speculative with major uncertainties and doubtful performance.
3-35 3.2.4.2 Reduction of uranium in molten salt to metal A reducing metal, such as lithium or calcium, might be added to give a complete reduction of uranium fluorides to uranium metal which would then be separated from the fuel salt. A number of serious concerns, including the potential for a nuclear criticality and the development of practical methods to collect the uranium product, were identified. Again, this technique is not recommended. 3.2.4.3 Dissolution and separation by solvent extraction Several options were identified that would separate uranium using an aqueous process. For all of these, the low solubility of the salt in aqueous solvents (Table 3.1, Sect. 3.1.4) results in a much larger aqueous solution, without an established basis that all components of the salt go into solution. To separate the uranium by solvent extraction, several feed adjustments would be required. These include the addition of aluminum nitrate to complex the fluoride, addition of an oxidant to put the uranium and plutonium in the (VI) valence state necessary for extraction, and other additives to achieve the required salting strengths and to minimize corrosion. While solvent extraction has been used for a wide variety of conditions, a development program would be required to establish the specific operating parameters for this system. There is no known operating solvent extraction system that could simply be adapted to process the salt solutions. Therefore, a relatively large, complex (and expensive) system would have to be built. 3.2.4.4 Dissolution and separation by ion exchange After complete dissolution of the MSRE fuel sail (with the difficulties addressed previously), an ion exchange process might be the most efficient and practical separation procedure for uranium and perhaps plutonium. Feed adjustments would be required to give the optimum conditions for metal separations. These are likely to include oxidation of all uranium to the U(VI) valence, adjustment of pH without precipitation ofU, Zr, or Be compounds, and possibly the addition of (NH 4) 2 SO4 or other salts to give anionic complexes ofUO 2* The ion exchange separation would probably be a batch operation with preferential removal of uranium and perhaps plutonium. The preferential removal of the main sail components (Zr, Be, and Li) leaving uranium in solution would require excessive volumes of ion exchange resin. Ion exchange separation of uranium is likely to produce a concentrated uranium solution product. Impurities in the product arc not significant if it is for the purpose of disposing the uranium. Losses of uranium to the larger volume of solution waste are only important if they affect prospects for disposal of this material. The used ion exchange resin itself might retain enough uranium or fission product activity to require disposal as a radioactive waste. An ion exchange separation of plutonium is also likely to be effective and practical. But the conditions for plutonium separation probably differ from those for uranium. Thus, separation of both uranium and plutonium would probably require two feed adjustments and two ion exchange treatments. 3.2.4.5 Dissolution and separation by precipitation The precipitation of uranium from the other materials in the MSRE salts has much different requirements than those that commonly apply. The principal requirement is separation from beryllium, lithium and zirconium salts in solution. Requirements for product purity and recovery percentage are less
3-36 restrictive for waste disposal than for the production of fuel. Impurities in the uranium precipitate are of little importance up to the point where they result in significantly larger amounts of waste solids for disposal. Losses ofuraniwn lo the waste solution are only important when they affect the prospects for disposal of that solution. The peroxide precipitation of uranium as U04*HP solids is more selective for uranium than the more commonly used ammoniwn diuranate or anunonium uranyl carbonate precipitations. The solubility of uranium oxalate is probably too high to give an acceptable recovery of uranium by oxalate precipitation. The separation from zirconiwn is probably the poorest and the controlling separation. The separation factors would have to be detcm1incd by experimental tests. After the salt was in solution, the pH should be adjusted upward to near 2, or as high as possible without precipitation of zirconium. After addition of peroxide and perhaps additional pH adjustment the solids and solution should be separated and analyzed. The uranium content of the solids and of the solution would be the critical results. Good yields and separations from impurities have been demonstrated but for significantly different solutions than would result from dissolving the MSRE salts. 3.2.4.6 Preferential dissolution of uranium from pulverized solid salt The solubility of U(VI) as U0 2 F2 is high compared to the solubilities of lithium, beryllium and zirconium fluorides. The concentration of uranium is much lower than that of the three major components. Therefore, a small volume of equilibrium solution could dissolve a high fraction of the uranium with only very small fractions of the other salts. Leaching of a soluble component from solids is a physical process dependent on the fineness of the solids and other conditions. Empirical, experimental infomrntion is required. If mechanical removal of solid salt is selected, then the usefulness of perfonning these tests should be considered. Leaching has no application if the salts are dissolved. Fluoride volatility is preferable for separation of uranium from molten salt. Some MSRE studies were made lo measure the aqueous solution concentration after fluoride salt powders were stirred in water. The solution concentrations showed higher U/Li and U/Be ratios than those in the salt. Oxidation ofU(IV) to U(VI) would greatly increase the solubility of uranium. The preferential leaching of uranium would give a U0 2 F2 and HF solution saturated with the other fluoride salts. Additional processing would be necessary to give uranium solids acceptable for storage. After the solution concentrations arc determined by experimental studiesof leaching, procedures for conversion to uranium solids could be selected and tested. Precipitation of U0 4*H 20 and calcination to U30 8 would probably be satisfactory. 3.2.5 Screening of Processing Technologies Fluorination is a technology for separating uraniwn from molten salt that has been used successfully in MSRE in 1968, is simple, and interfaces well with facilities that will have been constructed for the conversion of the UF6 present in the MSRE off-gas system to l!i Qi for stable, long-term storage and transfer to the fissile materials disposition program. Electrorefining is a technology that can separate uranium, transuranics and actinides, and fission products from the sail matrix. An eleclrorefiner based on simulated MSRE fluoride salt is now being demonstrated at Argonne National Laboratmy. This technology produces a range of product streams, and achieves both uranium separation and stabilization waste fom1s containing transuranics, actinides,
3-37 and fission products. Disposition of some of the waste products may require additional processing, combination of wastes from similar processes, or a waste qualification process. Other separations technologies have been identified, but either require large, expensive facilities for implementation or are technologically immature. None of the other technologies appear to offer important features not offered by fluorination or electrorefining. 3.3 STABILIZATION TECHNOLOGIES 3.3.1 Stabilization of Fluoride Salt with a Getter As discussed in Sect. 1.2.4, irradiation of solid salt liberates fluorine from the salt matrix. At elevated temperatures, the fluorine radicals produced back-react ,,~U1 the metal sites left behind, restoring the fundamental salt chemistry. However, at room temperature the fluorine radicals combine to fonn fluorine molecules, which diffuse through the salt and are released as fluorine gas. In addition, the fluorine released by radiolysis appears to oxidize uranium in the salt matrix, producing UF6
- This, too, diffuses out of the salt matrix. Preliminal)' experiments have suggested that fluorine is liberated from the salt matrix by beta and gamma, not alpha, radiation. Production rates have been quantified under a number of conditions, and projected future fluorine production is plotted in Fig. 1.16.
Generation ofiliese reactive gases and their migration have led to the most important issues now faced by the MSRE Remediation Project. Any package Urnt contains fluoride salt Urnt is exposed to a significant flux of beta-gamma radiation will require measures to control U1e evolution of fluorine and UF6* Reacting the gases wiili an appropriate chemical (referred to as gettering) is the simplest approach to controlling this phenomenon. This technique could be applied to control the production of both F2 and UF6 , or it could be applied to salt from which the uranium has been removed, in which case only fluorine is the issue. Since technologies are available for the separation of uranium from fluoride salt, and since the consequences of transport of uranium out of the salt matrix are potentially severe, gettering is considered a more appropriate technique for controlling fluorine only, except in cases where gas control may be required for a relatively short time. If needed, a sodium fluoride trapping system can be used to control UF6 evolution. The reactive gas removal system now being brought into operation is an implementation of this technique, and the trapping ofUF6 on NaF will not be discussed further. If the uranium is separated from the salt, and the salt residue is proposed as a waste fonn. a getter might be placed into the salt package. As seen in Fig. 1.16, the production of fluorine falls off as the sources of beta and gamma radiation in the salt decay away. Thus, the amount of getter needed for the fuel salt waste is that needed to ensure that -200 moles of fluorine are reacted. A getter can be used in two ways. A package of chemical can be placed in the container but as a discrete unit adjacent to the salt. In U1is case, fluorine is allowed to diffuse out of the salt matrix, travel to the getter, and react with the chemical in the getter. If the package is vented, it should be designed so that any gas leaving the package passes through the getter. Alternately, a material could be integrated into the salt matrix itself. In U1is case. the fluorine reacts with the getter before it diffuses out of the salt. As irradiation of the salt proceeds, isolated metal sites are formed at the locations where fluorine has been liberated. Eventually, these sites begin to serve as a getter. Experiments have demonstrated a "damage limit," at which point sufficient metal sites exist that fluorine gas evolution stops. This limit is reported to be about 2% fluorine loss. The loss of fluorine from the MSRE fuel salt is estimated to be
~ - - - - - - - - - - - ~--** --~~--------, ----
3-38 about an order of magnitude lower than this, and thus the damage limit will not be reached before the source decays away. A veiy preliminary review of getter materials has been pcrfonned (personal communication, D. F. Williams, July 25, 1996). The initial, conservative recommendation is that the use of a simple overpack of reactive solids would involve a reasonably straightforward laboratoiy demonstration. The number of equivalents necessary can be predicted by simple stoichiometiy. An initial proposal should recommend the use of proven fluorine getters such as soda-lime or alumina, even though active metals like aluminum or titanium sponge may have better performance. These other materials may be considered after some preliminary effectiveness tests have been performed, and the potential reactions with containers, other components of the gas, and the salt have been considered. Until performance tests are available, a stoichiometric excess (such as 200% of that required) will be assumed. A getter in an overpack should be separated by some distance from the salt in order to minimize the potential for re-rndiolysis of the fluoride salt in the overpack. The reactions that take place using soda-lime (5-20wt% NaOH. 6-I 8wt% water, balance CaO) are and the reaction for alumina is The gettering of the salt in a more intimate manner, either by alteration of the salt chemistiy on microscopic scale, or by blending of a second phase into the salt, is a for more challenging goal and requires additional development. It does, however, eliminate all issues associated with separation of salt and getter material or the interaction of materials with fluorine between salt and getter. Introduction of reducing sites into the salt might be accomplished by adding a soluble transition metal salt that has a stable lower oxidation state [e.g., Ce(Ill/IV)], by electrochemical means, or by addition of a second phase which is well distributed in the salt. If soda-lime (assumed to be 10% Na OH, I0% H2 0, and 80% CaO) were used in conjunction with the packages described earlier in this report, if only the CaO were considered in the calculation, and if twice the amount of CaO were provided as is necessaiy to react with the fluorine, then about 28 kg of soda lime would be required for all of the fuel salt. Each of the 24 packages would contain a little *over 1 kg of soda lime. Such a package will occupy a few liters of space, consistent with the volume of the package not occupied by salt (about 25%, or about 20 L). 3.3.2 Conversion of Fluoride Salt to Oxide Using the INEL New Waste Calcining Facility An alternate approach to stabilizing the fluoride salts is to convert the material from a fluoride to an oxide. An existing technology that could be applied is calcining, such as is perfom1cd at the INEL New Waste Calcining Facility. This facility uses a kerosene-fired spray calciner to solidify waste solutions from the tank fonn at the Idaho Chemical Processing Plant (ICPP). A general flowsheet of the INEL calciner is shown in Fig. 3.19. Wastes arc transported from the waste tanks, excess water is evaporated off, and the feed is injected into the calciner which operates at 500°C (932°F). The calcine particles accumulate in a fluidized bed, and arc drawn off at the bottom and pneumatically transported to the calcine storage bins. The off-gas (containing nitrogen oxides) is led to a cyclone to separate any
- 1.
3-39 Na-Bearing UquldWaste Storage Tank& Stac:Jc Venturi Scrub Quench Tank r ..K_o-~ Filtration Cyclone Pot Filter Feed ----~sooc Leach Surge Calcine Storage Bina Fig. 3.19. Diagram of the overall operation of the INEL New Waste Calcining Facility.
3-40 entrained calcine particles. It then goes to a series of off-gas treatment steps before being released to the stack. The feed to the calciner is an aqueous nitrate salt solution. The primaiy reactions taking place during calcination of typical INEL feed are (I) the decomposition of nitric acid; (2) the evaporation of water; (3) the conversion of the alwninum and stainless steel nitrates to the metal oxides; (4) the conversion of zirconium fluoride to zirconium oxide, and (5) the fonnation of zirconium fluoride (personal communication, P.A. Ostby, July 15, 1996). The fission product elements are converted to oxides if the nitrate is unstable at the calcination temperature. Cesium and strontium probably remain predominantly as nitrates, as does sodium nitrate and possibly calcium nitrate if it is present in excess of the amount that reacts with the fluoride in the feed. The simplified chemical reactions taking place in the calcination process are as follows: Metal nitrates - Metal nitrates/oxides , Fission product nitrates - Fission product nitrates/oxides. To prevent corrosion of the stainless steel off-gas equipment during calcination of fluoride-bearing waste, calcium is added (as calciun1 nitrate) before calcination to retain the fluoride in the calcined solids. The calcium reacts with fluoride in the feed solution to form a gelatinous precipitate of calcium fluorozirconate (CaZrFJ. This gelatinous precipitate remains suspended in the feed solution and during calcination at 500°C (932 °F) reacts to fom1 calcium fluoride, according to the reaction: The calcium fluoride is themrnlly stable at the calcination temperature of 500°C (932 °F). Calcination of the sodium-bearing waste, which contains up to 2 M sodium nitrate, is difficult because the solid sodium nitrate, which is molten at 500°C (932 °F), causes agglomerates to fonn in the fluidized bed of the calciner. To prevent agglomeration, sodium-bearing waste is blended during calcination with zirconium fluoride waste to fom1 ciyolite (NaAIF6). The resulting calcine is thennally stable at the calcination temperature. Typical product resulting from calcination consists of granular particles that are nearly spherical. The particle diameter of the product from the bed is generally in the range of 0.2 to 0.6 mm, but the diameter may vaiy from about 0.1 to 1.5 mm. It is free-flowing and readily transported pneumatically through pipelines. Most of the particles that leave the calciner vessel in the off-gas are removed by the primaiy cyclone and routed to the storage bins along with the caleiner bed material.
3-41 It is probably feasible to dissolve the MSRE fuel snit in an aqueous nitrate solution (with the difficulties discussed in Sect. 3.1.4) and blend the MSRE salt solution into feed to the INEL calciner (Denney N.d.). The corrosivity of the MSRE snit is not significantly different from that of other streams that can be handled (and have been handled) by the New Waste Calcining Facility. In particular, the beryllium content is no more troublesome than the cadmium content of other wastes. If the uranium is still present in the salt, the salts will be fed into the input stream of the calcincr gradually, with full regard for criticality issues. The resulting calcine would be stored in an existing facility instead of a new one yet to be built. The calcine resulting from the calcination process switches the fluorine content of the MSRE salt to CaF2, with sufficient calcium added to the mix to retain the fluorine. All of the fluorine that enters with the MSRE salt solution remains in the calcine. The product would be blended with the calcine already in storage at INEL. This option appears to be technically feasible, and makes use of an existing facility. It would require the restart or construction of a salt dissolver, and provision of adequate piping from these to the calciner feed. The chemistry of calcining the MSRE salts should be reviewed to ensure that no operational difficulties, such as those discussed for the sodium-bearing wastes, arc encountered. In the case of sodium-bearing wastes, the addition of zirconium provides a solution to operational difficulties in the calciner. It is not known whether lithium behaves similarly to sodium in the calciner feed. Both lithium and zirconium would be present in the MSRE feed (lithium in higher concentration than zirconium). 3.3.3 Conversion of Fluoride Salt to Borosilicate Glass 3.3.3.1 General approach to vitrification Vitrification involves combining a liquid, solid or slurry waste material with a glass-forming frit (typically SiOi, B:P3, and Na:P) at moderate-to-high temperatures [ l 050-1200°C ( 1922-2192 °F)] to produce a glass. The glass is then poured into a container and solidified. A typical high-level waste glass container is depicted in Fig. 3.20. Borosilicate glass is the only waste fonn for which Preliminary Waste Acceptance Specifications for disposal in the federal repository exist. A borosilicate glass process has been used in France by COGEMA to immobilize high-level waste slurries for a number of years. Vitrification plants have recently started up at the Savannah River and West Valley sites. A proposal has been developed for the construction of a vitrification facility at INEL. Bids are being evaluated for the demonstration of high-level waste management at the Hanford site, which could include vitrification. Use of the existing vitrification facilities, and the proposed INEL facility (which could immobilize calcine produced at INEL) are discussed in the following sections. 3.3.3.2 Conversion of fluoride salt to borosilicate glass using the Savannah River Defense Waste Processing Facility Weapons material production at the Savannah River Site (SRS) has produced unusable byproducts such as intensely radioactive waste. The high-level radioactive waste, about 35 million gallons, is stored in waste tanks on site. The S-aren DWPF "~II bond the radioactive elements in borosilicate glass, a stable storage form (WSRC-RP-94-396 N.d.). Figure 3.21 shows a schematic diagram of the process to immobilize high level waste stored at SRS. The DWPF sludge feed slurry is considered to be well-mixed because of the extended sludge processing in the tank fann.
3-42 I I
---~*s"* I*--
I 1650 kg WASTE 10' GLASS 7'6" 470 kg CANISTER (EMPTY) r-2*-. Fig. 3.20. A typical borosilicate glass high level waste package.
3-43
- - - - - - Cemelll & Ay..:
Salt Sohb. 'G SOLUBLE SALT ILUI>GE SALT PROCESSING 1-'onnJc SLURRY CELL Add RECE ll'T& ADJUSTMF.NT I INCINERATOR Nitric Add TANK FRIT OFF GAS 0 CLEANING SLURRY MIX l:.'\'Al' ORATO R CANISTER CLOSURE DECON INTERll\1 STORAGE FOR E\'ENTl!AL SIIIPMEh'T TO A FEDERAL REPOSrtORY
,,:_::.,-$.,*.~;~"-;--*
~[;i~~*?I Ml-:J.TER 1"1-:El>
TANK MELTER Fig. 3.21. Diag ram of the overall operations perfo rmed at the Savannah River Defense Wast e Proc Facility. essing
3-44 Waste salt solution is processed in the in-tank precipitation process to remove 137Cs and to adsorb strontium on monosodium titanate. Soluble uraniwn and plutonium are also adsorbed on monosodium titanate. The tetraphenylborate slurry containing potassium and cesium tetraphenylborate precipitates and monosod.ium titanate with adsorbed strontiwn, uranium, and plutonium is then washed in the DWPF late wash facility before it is sent to the DWPF salt processing cell. In the salt processing cell, the precipitate sluny is first stored in the precipitate reactor feed tank and then processed in the precipitate reactor. In the precipitate reactor, tetraphenylborate is destroyed by acid hydrolysis reactions to produce benzene and an aqueous phase containing radionuclides. The sludge feed is slurried, washed, and filtered several times in Tanks 40, 42, and 51 and becomes homogenized during extended sludge processing operation. Alkaline sludge feed is then transferred from the tank farm to the sludge receipt and adjustment tank in the DWPF chemical processing cell. The sludge sluny is treated with nitric acid in the sludge receipt and adjustment tank to control the rheological properties of the sludge slurry, primarily nitrites, carbonates, and soluble hydroxides. The precipitate hydrolysis aqueous product is combined with the sludge sluny and the borosilicate glass :fiit in the DWPF chemical processing cell to produce melter feed. The high level radioactive waste is finally immobilized in a glass matrix contained in sealed stainless steel canisters. To summarize the overall DWPF process operation, there are two high-level waste feed streams from the tank fam1 delivered lo the S-area DWPF. The precipitate feed sluny is processed in the salt processing cell to produce the precipitate hydrolysis aqueous product for subsequent sludge receipt and adjustment tank and melter operations. The sludge slurry is processed in the sludge receipt and adjustment tank, and combined with the precipitate hydrolysis aqueous product. The amount of precipitate hydrolysis aqueous product, or more correctly the amount of soluble and insoluble solids in the product, is half of the sludge solids. All of the waste sluny feeds are combined in the slurry mix evaporator with the glass fanning friL The melter immobilizes all of the waste feeds in a waste borosilicate form. It has been proposed that th_e MSRE salts be dissolved and blended into the feed to the DWPF. The acceptability determination for DWPF process control (Brown and Postles N.d.) limits the fluoride content in the DWPF feed to I g NaF per I 00 g glass. Assuming this limit is based on the effect the
- fluoride has on the process, a molar conversion was performed to define the rate al which the MSRE salt
- fluorides could be bled into the DWPF feed stream, and the number of glass canisters in which that the MSRE material would be spread. These calculations (personal communication, A. Dubbs, July 30, 1996), based on information supplied by Savannah River personnel (personal communication from T.
Gutman to A. Dubbs, July 17, I 996), indicate that the MSRE fuel and flush salt each would be distributed through at least 500 waste canisters. If there is already a significant fluoride content in the Savannah River feed, this number would increase. Thus, the MSRE blending operation would require at least 25% of the total DWPF operating campaign to immobilize the contents of the SRS tank farm. The feed to the DWPF would be an acidic solution. Again, this would require either converting existing equipment in the Savannah River canyons to this purpose or constructing a new dissolver. Waste form performance is an empirical art, and lest melts would have to be evaluated to ensure that an acceptable waste form results from the proposed blend. No technical reasons have been identified to date that rule out encapsulation of the MSRE salts in the DWPF, but it does not appear to be a short-term operation to be conducted al minimal cost.
3-45 Other approaches to vitrifying materials in the DWPF or a similar facility have been proposed by Savannah River personnel. These include preparing melts of higher concentration material that is then placed inside a standard canister, and surrounded with standard melt. Development of these techniques is still at an early stage. 3.3.3.3 Conversion of fluoride salt or calcine to borosilicate glass using the proposed INEL Waste Immobilization Facility The INEL has been converting liquid high-level wastes to calcine for a number of years. However, the calcine is considered to be a material suitable for interim storage but not for disposal in the federal repository. Thus, a proposal has been developed for a waste immobilization facility to convert the calcine (and any liquid waste that would remain in the tanks) into a waste fonn that would be accepted in the repository (Denney N.d.). Borosilicate glass and glass ceramics fonned in a hot isostatic press are the candidate waste fonns. This discussion focusses on the borosilicate glass waste fonn. The INEL waste immobilization facility has not yet been approved. However, the settlement agreement stipulates that high-level wastes leave Idaho by 2035 (Settlement agreement N.d.). Design of the waste immobilization facility is expected to begin around the year 20 l 0. The overall steps of the proposed waste immobilization facility are shown on Fig. 3.22. The first phase of the project would address any wastes that remain in the liquid waste storage tanks, and later would treat dissolved calcine. The proposal treats the waste liquid in a number of stages to concentrate actinides and fission products for immobilization in the vitrification facility. The low activity liquid waste that remains would be solidified with grout and would ultimately be disposed of as low-level radioactive waste. Actinides would be separated from the liquid waste using the TRUEX continuous counter-current solvent extraction process. A set of centrifugal contactors will be used to remove an aqueous concentrated actinide product and to send the raffinate to the next process step. The second process step is based on a similar solvent extraction system, using a crown ether (such as 18-crown-6) to remove strontium. Again, the product is a concentrated aqueous stream, and aqueous raffinate is sent to the next process step. The final process step is the removal of 137Cs by ion exchange. Either spent resin or an eluant solution becomes the final high-level waste stream, and the effluent is sent to the grouting facility. The concentrated aqueous actinide and strontium wastes and the concentrated cesium wastes from the ion exchange process are then recombined and become the feed for the vitrification facility. The second phase of the project will be the construction of the vitrification facility and a calcine dissolver. Dissolved calcine will be sent through the separations process described above. The vitrification flowsheet is shown in Fig. 3.23. For a liquid feed, the waste solution would be sampled* and analyzed, and based on the analytical results, a suitable frit would be selected. The liquid would be dried in a rotary cnlciner or blended directly with frit and introduced to the glass melter as a sluny. If calcine were to be vitrified directly, it would be mixed with glass frit in the required proportions and fed to the glass melter. Typical calcine waste loadings in the glass vary from 25 to 33 wt% depending on the type of calcine to be vitrified. The glass melter would typically be operated at a temperature between 1050-1200°C (1922-2192°F). Molten glass produced in the melter would be poured into canisters, sealed, and transported to an interim storage facility before final disposal. The melter off-gas consisting of hazardous (Hg and Cd) and radioactive (Cs, Tc. and Ru) materials would be treated in an off-gas treatment facility and recycled to the melter as needed before stack discharge.
3-46 p--------------------------------------------1 i§"""~=seJ)arano - I -*
--~=
I
-= acili
/~*
*>;*:~-~~:><-~~
,* . . *"/** ***
. : ~
1-=U"'A-=:t'fT=A:.ccum-
~7~iffificatio -
,~w *ulatiOD
.:*.**Dissolver;;.-:::- ..-:/
_..-~-- ,:-:*'**** .:-:** .-.: lL[lJJWIU.On --**. :
*** .-*.-. ., . v::.-:-J:,.,...:* * )I: HAWT-*- -=..:.*
.. ,:** .. iim.uw
*.-*._ ....... _:*** *:**** :*:>pJant .:*. **
- _.-.* . >storag*eF',.,.mm
. . ~ . Qlw~J **
g
*:... ':-* * ** * ,I' * ..- ... :.. * :- ** **
- i" .
r----- c -"-.*** --- **::J Phas e*I ._______1 Phase II Fig. 3.22. Diagram of the overall operations that would be performed at the proposed INEL waste immobilization facility.
3-47 Additives B Ii
,-.--1 a Filtration r.u Condenser LiquidHLW Storage Tanks GRcmcval Calcine Storage Bins I ~cm~C alcine Blender Glass Waste Canister Fig. 3.23. Diagram of the overall operations that would be performed at the propose d INEL vitrification facility.
3-48 This process is similar to those implemented at Savannah River, West Valley, and by COGEMA in France. Many features of these facilities are expected to be directly applicable to the vitrification of INEL wastes, reducing the amount of research and development needed to implement the INEL process. Nonradioactive, pilot scale tests of joule-heated ceramic glass melters completed at the ICPP on simulated waste have proved the initial feasibility of the calcine vitrification flowsheet. It is expected that the MSRE salts could be vitrified in this proposed facility. The MSRE material could be fed as a component of the INEL calcine, or it could possibly be fed directly as a solid or acidic solution. Because of the preliminary nature of the description of the INEL facility, no details on the requirements for accommodating the MSRE materials are available. 3.3.3.4 Conversion of fluoride salt to borosilicate glass using the West Valley Demonstration Project Vitrification Facility A vitrification facility, in many ways similar to the Savannah River DWPF, has recently begun operation at West Valley, New York, to immobilize wastes remaining from the operation of a commercial reprocessing facility in borosilicate glass. It is assumed that the operation of that facility is limited to the wastes already present at tl1e West Valley site, and use of this particular facility has not been explored. The general technical issues identified with the use of the DWPF would likely apply to the West Valley facility as well. The waste volume in which to dilute the MSRE material would be less, and the facility would be even less able to adapt to an unusual inlet stream. 3.3.4 Other Salt Conversion Processes 3.3.4.1 Oxidation and conversion of fluoride salt to glass using the Glass Material Oxidation and Dissolution System (GMODS) The GMODS process has been developed to vitrify a wide variety of materials that do not exist in the oxide form. Existing vitrification processes convert only oxides or oxide-like materials to glass. GMODS has been developed to directly convert oxides, metals, ceramics, organics, halogens, and amorphus solids to glass. This can allow the conversion of complex waste mixtures, including filters, process wastes, laboratory wastes, etc.) to glass without pre-processing. GMODS has also been considered for the direct conversion of halogenated materials (initially chlorides, but recently the MSRE fluorides) to glass (Forsberg and Beahm 1996). The GMODS process converts the feed materials to the glass form inside the melter. The process can operate in batch mode (as shown in Fig. 3.24), or in continuous mode. This discussion describes the batch process. The process begins with a glass melter filled with molten, lead-borate dissolution glass. Oxides dissolve in glass, but metals and organics do not. The GM ODS process uses lead oxide (PbO) in the molten glass to oxidize metals to metal oxides and organics to carbon oxides. The resultant metal oxides dissolve into the glass. The carbon oxides exit the melter as gasses. The lead metal reaction product separates from the glass and fom1S a separate layer at the bottom of the melter. The boron oxide (Bi0:J in the melt assures rapid dissolution into the glass of any protective oxide layers on metal wastes. After dissolution of the waste, silicon oxide and other additives are added to the glass to produce a high-quality product glass. Excess PbO is removed from the glass by adding carbon, which converts the PbO to lead metal and carbon dioxide (CO2). The final glass may have some or no PbO depending on the desired product glass. The product glass is poured from the melter into the waste packages. To generate the next batch of dissolution glass, boron oxide is added to the melter and the lead metal is oxidized to PbO with oxygen.
ORNL DWG 9 A-97 R Glass Fritl Waste Carbon a. Glass Glass Glass Lead Lead Lead LEAD BORATE
- ADD WASTE
- ADD GLASS ADDITIVES DISSOLUTION GLASS
- OXIDIZE/DISSOLVE WASTE
- ADD CARBON TO REDUCE EXCESS PbO IN GLASS
'.I Boron Oxide 1
- l,J I
i 1 Glass .i::,. I, Lead Oxide * \0 (optiona/J Lead Glass Lead Oxyge n---- Product Cylinder
- ADD MAKEUP 82 03
- ADD OXYGEN l.
- ADD MAKEUP PbO
- OXIDIZE LEAD TO PbO POUR GLASS Fig. 3.24. Process steps for batch processing using the glass materials oxidation and dissolution system.
3-50 GMODS can convert halogen-containing materials to glass, a process which creates a separate nonradioactive sodium halogen waste stream. Halogens, such as fluorides, usually make poor-qua lity glasses. In the GM ODS dissolution glass, fluorides in the waste produce lead fluoride (PbF ), which 2 is volatile at glass melter temperatures and exists to an aqueous sodium hydroxide (NaOH) scrubber. In the scrubber, the PbF2 will react with the NaOH to yield insoluble lead hydroxide [Pb(OH}J and soluble NaF. The insoluble [Pb(OH)2] is recycled back to the melter, where it decomposes to PbO and steam. The aqueous NaF stream is cleaned and discharged as a chemical waste. Because of the corrosive characteristics of the initial dissolution glass, GM ODS requires a cold-wall melter in which cooling jackets in the walls produce a "sh.-ull" of solidified material that protects the walls from the contents of the melter. The melters can be heated by fossil, induction, plasma arc, or electron-beam systems. Such systems are currently used to melt high-temperature materials (e.g., titanium and superalloys) and to produce specialty glasses. The basic GMODS process creates a lead borate fusion melt. From that melt a borosilicate glass can be produced. Alternatively, the lead oxide could be removed from the melt by adding carbon, producing a borate fusion melt. This material could then serve as a feed to a standard borosilicate glass melter, such as DWPF. Tests perfonned to date have demonstrated the dissolution of U0 , Zr0 , Al 0 , Cep , MgO, and 2 2 2 other oxides. Oxidation-dissolution tests demonstrated the oxidation of uranium, 3cerium,3zircalloy-2, aluminum, stainless steel, and other metals. Oxidation-dissolution tests have also demonstrated the oxidation of carbon and graphite, with production of CO
- The other process steps (adding glass 2 frit, removing lead from the glass, and oxidizing lead back to PbO) have been investigated in the laborator y.
The reactions necessary to convert the materials in the MSRE salts have been reviewed using literature data, but have not been verified in the laboratory. To move forward with application of GMODS to the MSRE salts, small laboratory tests are needed with lithium fluoride and beryllium fluoride to experimentally confinn the thennodynamic analysis of the behavior of fuel salts in the GMODS process. Discussions should be held with DWPF personne l on the option offeeding a GMODS product into the DWPF. Design and construction of a GM ODS facility might then proceed. It is possible that a number of potential users can jointly fund the further development of a pilot-scale GM ODS facility. 3.3.4.2 Convers ion of fluoride salt to mineral phospha te or phospha te glass An alternate proposal for conversion of the MSRE fluoride salts involves the production of a mineral phosphate or phosphate glass waste fonn by reacting the salts with boron phosphate (personal communication, C. Baumberger and D. Beach, May 16, 1995). This proposal is based on the reaction of boron phosphate (BP0 ) with the molten (or granulate 4 d solid) salt mixture to remove the fluoride ions from the system and to generate highly stable metal phosphates. The basic reactions arc shown below for the major constituents of the salt:
3-51 The driving force for these reactions is the formation of boron trifluoride (the bond dissociation energy ofBF3 is 145 kcal/mole, indicating a very strong boron/fluorine bond) which is evolved as a gas, further driving the reaction Lo completion. The reaction could be carried out on the cold, solid salt, should some method of removing the salt from the vessels as a solid be used, or the reaction could be carried out on the molten salt removed from the vessels. The boron trifluoride would be collected, hydrolyzed, and disposed of as very low-level waste. Because the process uses solid-solid (in the case of solid salt) or solid-liquid (in the case of molten salt) reactions, no solvent would be required, minimizing any waste disposal problems. It may not be necessary to remove the uranium as it too would be converted to a highly stable phosphate complex, but it might be necessary to dil~te the phosphate products in a glass to remain under limits for fissile material concentration or mass in a waste package. If the uranium were removed by fluorination, this would not be an issue. The lanthanum orthophosphates are analogs of the mineral monazite. This mineral is exceedingly stable, does not readily leach, and is resistant to alpha-particle damage in naturally occurring thorium-and uranium-containing materials. In addition, a study of the use of phosphate glasses as an alternative to borosilicate glasses shown superior resistance to leaching. Lithium orthophosphate per se is one of the least soluble salts of lithium. To advance this proposal, experiments are needed to obtain the identify of the phosphates and the kinetics of the reaction of boron phosphate with a laboratory mixture of the fluoride salts present in the MSRE. While the individual reactions have been shown in initial tests to proceed to completion at temperatures above I000 °C ( 1832 °F), it is necessary to determine of double or triple phosphates may form, and it is desirable to obtain information on the kinetics of the reaction with mixtures of salts. There is a belief that the reaction in molten salt may proceed considerably faster than the solid-state reactions of the initial tests. This is due to the fact that reactions in the solid-state often require high temperature and/or prolonged reaction times because of the slow diffusion of reactive species in the solid state. The reaction in molten salt should not suffer this problem. This development program would provide engineering data for the design of a conversion system. It would also make recommendations on the materials of construction, optimum temperatures, and method of introduction of the reactants. 3.3.5 Conversion of Uranium to U3 O8 3,3.5.1 Conversio n of UF6 to U3 O8 Conversion oflJF6 to U3 O8 is being addressed by the MSRE Remediation Project as part of the campaign to place the uraniun1 now present in the off-gas system into safe storage in the ORNL RDF. Initially, the uranium will be stored as UF6 trapped on NaF. A commitment has been made to convert this material to oxide within three years of being placed in temporary storage on NaF. A study into alternative flowsheets for conversion of UF6 to oxide has been perfom1ed (Advanced Integrated Management Services 1995). This report will only summarize a likely selection from that study, and will not attempt to address all of the options considered. The flowsheet for one of the conversion processes recommended to convert the uranium collected from the off-gas system is shown in Fig. 3.25 (Sherrill 1995). The uranium enters the process chemi-sorbed on a NaF trap. The UF6 is volatilized by heating the NaF in the presence of a fluorine gas stream. The fluorine, fed as a 5% mixture with heliwn, maintains the uranium as UF as it passes to a cold trap. 6 Solid UF6 is condensed out of the outlet gas from the NaF in a cold trap. The other gasses then pass through a NaF trap for the removal of any uranium carried over from the cold trap, and an activated
N2 1110 N2 S%F 1110 Ni Chemical (l/-23R) 2 spa,ge addilion FE Process off-gas To Refrigeration
~ Gp System I
Calchrntinn llot Cell ,O I Vessel with I UO, lo Packaging
!liid Storage I
I Rupture Disk
- - I I
NaF Activated I w Trap Alumina I I V, Trap N Surge --@--!I Tank l>esorptlon I 'i Furnace I J
! Vacuum 1'11111p Precipitation and Calcinatlnu l'urnace Liquid Waste To 1.1.1.W Tank
( 'nnJcnsatc Smi,plc snlnple Fig. 3.25. Proposed overall flowsheet for the conversion ofUF6 to oxide.
3-53 alwnina trap to remove the excess fluorine. The inert gas stream is then discharged into the process off-gas system. The UF6 is then hyrolyzed to UO:?2 with water in the cold trap, and flushed over to the precipitation and calcination furnace vessel. Uranium is precipitated either as ammonium diuranate by the addition of ammonia, or as a peroxide. The solids in the vessel are then washed with water, and then heated to drive off excess water. They are then thermally calcined at temperatures ranging from about 575 °C to 800°C (1067 to 1472°F). The resulting oxide is removed for analysis and packaging, and is sealed in a storage container that conforms to the design of the ORNL RDF handling systems. 3.3.5.2 Conversion of uranium metal to U3 O8 An alternative to fluorination, electrorefining, can produce a metal uranium product. Uranium is a reactive metal. If this material is to be placed in storage in the ORNL RDF along with the rest of the 233 U inventory, it is preferred that it be converted to a stable oxide. Specific flowsheets for the conversion of uranium metal to oxide have not been addressed to date as part of this effort. If the electrorefining approach is selected, with the transfer of uranium to the ORNL RDF as oxide, a flowsheet for this conversion process will be identified. Uranium conversion processes have generally been studied by the nuclear fuel and weapons industries, and identification of a flowsheet for this conversion is not considered to pose a risk to the implementation of the electrorefining alternative. 3.3.6 Screening of Stabilization Technologies Even with the separation of uranium, further stabilization of the salt is necessary to prevent the release of fluorine produced by radiolysis of solid salt. The simplest approach is to react fluorine with a getter, such as soda lime, as it forms in the vapor space above the salt. This provides a waste form that can meet the acceptance criteria for WIPP. It may not, however, be certified as a waste form for the federal repository, which relies more strongly on the performance of the waste matrix to achieve its repository performance goals. Packaging the salt \\'ith a getter does not preclude the possibility of further processing into an alternate waste form at a later time. The electrorefining technology discussed in .the previous section can function as a stabilization technology as well as a separations technology. As proposed by Argonne National Laboratory, 'the electrorefining technology produces a zirconium/noble metal waste, a waste consisting of actinides and fission products encapsulated in metallic bismuth, a metallic uranium product, a metallic waste encapsulating the 90Sr fission product, and a chloride waste salt contaminated with plutonium and other radionuclides. The disposal of the zirconium metal waste and the chloride salt is linked to waste management activities associated with the processing ofEBR-II wastes at ANL-W. This technology is currently being demonstrated with simulated MSRE fuel salt at Argonne National Laboratory-East. It requires a moderate-sized facility that might be installed in an inert atmosphere enclosure in an existing facility. Qualification of the bismuth metal waste fonn has not yet taken place. If the electrorefining technology were used only in a sequence that removes the zirconium metal waste, all actinides, transuranics, and most of the fission products in the bismuth waste form, and leaves the 90Sr in the salt for decay, a simpler process results. The metal waste form from this process might be further stabilized in a vitrification facility.
3-54 Vitrification of the salts, or the key radionuclides in the salt, produces the borosilicate glass waste fonn that has been deemed acceptable for disposal in the federal repository. Vitrification of the salt is technically challenging because of the high fluoride content. A test program of glass containing higher fluoride concentrations may be necessary to evaluate U1e resulting waste product. The fluoride salts may be converted to oxides by calcining at high temperature. The chemistry of calcining may need to be demonstrated for the materials present in the MSRE salts. The calcine stored at INEL is not considered a pennanent waste fonn. Calcine may produce a stable, although water-soluble, solid for interim storage. Dissolved calcine may later serve as a feed for vitrification. Other salt stabilization technologies, such as conversion to phosphate-based waste fonns or conversion in a glass material oxidation and dissolution system are interesting but are not technically mature at this time, and no actual conversion facilities exist. Technologies to convert uranium to oxide for storage in the ORNL RDF are generally well understood. A facility to convert UF6 to U3 O8 is expected to be available by the time the salts can be removed from the drain tanks. 3.4 PACKAGINGAND TRANSPORTATI ON 3.4.1 Salt Packaging and Transportation A proposed container for salt removed from the drain tanks is shown in Fig. 2.5. This container is consistent with packaging defined in the WIPP waste acceptance criteria, and presumably could be transported in the 72-B cask. The dose at the surface of the salt container, with two inches of steel as a shield, would be about IO R/h (Table 2.2). Without the internal shielding, the surface dose would be about 270 R/h. Although the weights and shielding requirements were not defined, it is presumed that these doses are consistent with use of the 72-B cask. If fluoride salt containing the gamma source from cesium or the 232U daughter chain is to be transported, evolution ofradiolytic fluorine must be controlled. A getter, such as soda-lime, is proposed for this purpose. The containers can be vented through the getter and a HEPA filter, if necessary. The licensability of such a transportation package needs to be reviewed. The amount of fluorine expected can be calculated and verified in the laboratory; thus the amount of getter required can be clearly established. The use of the 72-B cask is part of an overall transportation system being developed to ship wastes from a number of DOE sites to WIPP, including the shipment ofRH-TRU from ORNL to WIPP. The WIPP transportation progran1 has identified proposed shipment routes, and has been interacting with the states through which wastes pass. The technology could be used to transfer salt to other DOE sites (e.g., Savannah River or INEUANL-W). Definition of transportation routes and coordination of the sites may begin with the review of other shipments of radioactive materials between those sites. Transportation, particularly arrangements between states for acceptance of waste shipments, could significantly affect tlle implementability of several options. Shipping packages or casks will be needed to implement any of tl1e ultimate disposition strategies. The actual status of the 72-B cask, or other containers and casks that might be used to ship salt or salt components, has not yet been explored.
3-55 3.4.2 Uranium Packaging Uranium storage as an oxide at the ORNL RDF is practiced today. A container that was developed for the storage of oxide produced by CEUSP is readily adaptable for use by the MSRE project, and is being implemented in the planning and design of the facility that will convert uranium withdrawn from the MSRE off-gas system to oxide. Traps and shielded carriers have been designed and fabricated for the transfer ofUF6 on NaF within ORNL as part of the reactive gas removal activity. Packaging and transportation of uranium that might be separated at other sites, or that might be in other physical fonns, will be addressed if it becomes clear that such a strategy is desirable. 3.4.3 Packaging and Transportation of Other Waste Forms Most of the facilities that are vitrifying high-level waste are placing the waste glass in containers of similar dimensions, as depicted in Fig. 3.20. A transportation system will be developed for shipments of waste glass from the major sites producing such material to the federal repository. This system would address the MSRE requirements if a glass waste fonn is produced. Other wastes could be produced by processes such as electrorefining. The wastes produced by the electrorefiner are generally stable and physically small. It may be possible to transport those wastes in existing casks used to transport radioisotopes. Further definition of packaging and transportation requirements for these processes will be developed if the associated alternative is pursued further. 3.4.4 Licensing, Regulatory, and Other Issues Specific activities to address the licensing of shipments of MSRE materials have not yet been initiated. Licensed casks will be needed for transportation on public roads. Interfaces with the destination state, and other states through which materials pass, will be established once the content of specific shipments is identified. Of course, concurrence by the receiving state will address the overall activities to be conducted within that state's boundaries as well as the transportation issues themselves.
4-1
- 4. IDENTIFICATION OF ALTERNATIVES The alternative identification process described in Sect. 1.5 is based on achievement of the end points defined in Sect. 2. Seven principal alternatives are defined here. Some of the defined alternatives address in general a range of treatment options that can be evaluated collectively.
The alternatives that are selected for evaluation are based on the use of existing facilities, or on facilities that are currently being demonstrated with MS RE-type salt. Other, generally similar alternatives could also be proposed. An example would be to develop a glass form that performs adequately with a high fluoride concentration, and construct a vitrification facility at ORNL or another site. In several cases, alternatives might be applied to salt containing the uranium, or to salt from which uranium has been separated. The selection of representative scenarios using existing and planned facilities at the various DOE sites is based on the recommendation of the Programmatic EIS for DOE spent nuclear fuel. In fact, several of the options could be assessed either with or without the presence of the uranium. The first six alternatives evaluate scenarios for the ultimate disposition of the key contaminants in the MSRE fuel and flush salts. The seventh alternative addresses the case in which a commitment to a specific ultimate disposition schedule cannot be made at this time, and ends with a period of interim storage. A representative interim storage mode is selected for the initial evaluation; this is followed by a review of credible approaches to interim storage. It is presumed that uranium removal is necessruy to meet the WIPP acceptance criteria for 239Pu fissile gram equivalents. The effects of downblending the uranium to reduce the potential for criticality have not been considered at this time. 4.1 PERMANENT DISPOSAL IN THE DRAIN TANKS Permanent disposal in the drain tanks could take place by leaving the material in its present condition or by performing a number of activities to improve safety and protection of human health and the environment. In order to provide a baseline evaluation of other alternatives, "no action" is evaluated as an alternative. In this alternative, the salts are left in their respective drain tanks with no further actions taken. All facility operations are eventually assumed to cease. A number of enhancements can be applied to the storage of the salts in the drain tank. The most obvious of these is to monitor and operate the facility (including the sump pumps) and to operate the reactive gas removal system to prevent the further transport of uranium away from the salt. Additional enhancements could include the placement of neutron absorbers in or around the tank, placing a getter in the tank, filling the cell with grout, or even removing the salts, removing uranium and other components, and replacing the salt in the tanks. The enhanced storage alternative evaluated is based on facility operation and operation of the reactive gas removal system. This addresses near-term storage effectively, but breaks down eventually. For the long life of the 233U system, in a location below the natural water table, prevention of criticality by adding material external to the salt is problematic. If the salts are removed for processing, there is no incentive to place them back in the drain tank.
4-2 Thus, the alternatives for pern1anent disposal in the drain tanks are defined as follows: Alternative I, No Action. Continue to store the MSRE fuel and flush salt in their respective drain tanks, in their current condition. Assume all facility operations eventually cease. Alternative 2, Enhanced Storage. Continue to store the MSRE fuel and flush salts in their respective drain tanks, but implement and operate enhancements to control reactive gases, prevent nuclear criticality, and contain radioactive materials. These alternatives are depicted in Fig. 4.1. 4.2 DISPOSAL OF ALL KEY CONTAMINANTS IN THE FEDERAL REPOSITORY This approach places all of the significant fissile and radioactive material in the federal repository. It is these materials, and not the chemical carrier salt, that are of importance. Separation of uranium (the bulk of the fissile material) from other material is not absolutely necessary. Since there are existing technologies and facilities that can be considered for implementing this approach, the development of new technologies is not evaluated. The fuel salt is currently listed on the DOE spent fuel data base, and is referred to in the Programmatic EIS for DOE spent fuel. Because the EIS directs that non-aluminum-clad fuels are to be managed at INEL, the strategy developed by INEL for managing the salt is used to define this alternative. Options include storage of stabilized salt, calcining, and possible vitrification at a later date, before disposal in the national repository. The prinmry recommendation from INEL is to store as a calcine until a proposed immobilization facility (likely based on vitrification) becomes available. The alternative evaluated is thus defined as follows: Alternative 3. Dispose ofthe Salt, Including Uranium, in the Federal Repository. Remove the MSRE fuel and flush salts from their drain tanks. Stabilize the salts as appropriate for shipment. Ship the salts to INEL as spent fuel, in accordance with the Programmatic EIS for DOE spent nuclear fuel. Convert the salt to oxide in the INEL waste calcincr. Store the calcine at JNEL until the proposed remote-handled immobilization facility becomes available. Vitrify the calcine in the remote-handled immobilization facility. Ship the vitrified waste to the federal repository for ultimate disposal. Alternative 3 is shown as a flowchart in Fig. 4.2. Other technologies could also be used to implement this strategy. The salt, with uranium, might be blended into the feed for the Savannal1 River DWPF. The electrorefining technology developed by Argonne National Laboratory might be applied without the separation of uranium. (In fact, if uranium separation were deleted and an approach to vitrifying metal wastes from the electrorefiner were identified, a significantly more streamlined flowchart for the electrorefiner approach can be generated.) Since these technologies are evaluated for the approaches that do incorporate separation of uranium, they are not re-evaluated as part of this alternative.
J
- 1
'1
-~ ,i Salt, as-is Salt, as-is ~ ... ~ t' Run reactive gas removal, ~:,. sump pumps .i:,. l.,
- l!
- ~
Fig. 4.1. Flowcharts for alternative 1, no action, and alternative 2, enhanced storage.
- i
Salt, as-is Purge Store as and trap calcine Melt and
- Implement "',. Pack as *, Ship to "' Dissolve .... ,.. Vitrify withdraw ' gas control spent fuel INEL ,. and calcine calcine
.i:.
... ir J:.
Ship to repository Place in repository Fig. 4.2. Flowchart for alternative 3, dispose of the salt, including uranium, in the federal repository.
4-5 4.3 DISPOSAL OF SEPARATED URANIUM In assessing the hazards either in the near term, or in the vel)' long term (repositol)' time frames of 10,000 years or later), the hazards posed by 232U and mu are the major, if not totally dominant (in the repositol)'), fraction of the total hazard. For example, at 10,000 years of decay, the radionuclide inventol)' in salt residue from which uranium has been separated by fluorination would be about 46 Ci, whereas the radionuclide inventory of the uranium package that had been separated from the salt is about 1,800 Ci. Thus, strategies that separate the uranium from the rest of the materials in the fuel and flush salts are proposed. Technologies are available that can separate uranium from molten salt as either UF6 or metal. There are known technologies to convert these materials to oxide, such as U3O8
- The MSRE Remediation Project is committed to build a facility to convert the UF6 removed from the off-gas system to U3O8, and this facility should be available to convert UF6 removed from the fuel and flush salts as well. This oxide is the preferred form for storage in the designated national mu repositol)' in the Radiochemical Development Facility (Building 3019) at ORNL. A subset of the DOE Fissile Materials Disposition program will address the disposal of that uranium. Downblending and vitrification is one of the proposals under evaluation. If downblending is part of the strategy of the materials disposition program, it could readily be performed by adding depleted or natural UF6 to the MSRE 233 U~ fed into the conversion process.
Since the DOE Fissile Materials Disposition program is addressing the disposal of an existing (and much larger) 233U inventol)' of the same physical and chemical form, there is no incentive to address the disposition of uranium from the MSRE separately. The latter would only incur additional cost to DOE to provide two separate programs for the disposition of the same material. Thus, all alternative strategies that involve separation of uranium are based on conversion to oxide and transfer to the materials disposition program. This is reflected in the definition of alternatives in several of the following sections. 4.4 DISPOSAL OF THE KEY CONTAMINANTS IN THE SALT RESIDUE IN WIPP The fissile and radioactive materials in the MSRE salts include long-lived transuranic and actinide isotopes that warrant some fom1 of geologic disposal. DOE defense transuranic wastes are to be disposed ofin the Waste Isolation Pilot Plant. A review of the quantitative waste acceptance criteria for remote-handled transuranic demonstrates that these criteria can be met by fluorinated fuel salt that has been stabilized with respect to the generation of fluorine. Additional study into the acceptability of gettering and the potential risk associated with fluorine in WIPP are needed, but at this time no need for further processing of the salt is identified. This alternative, charted in Fig. 4.3, is defined as follows: Alternative 4. Transfer the Uranium 10 lhe Malerials Disposition Program and Dispose ofthe Salt Residue in WJPP. Remove the MSRE fuel and flush salts from their drain tanks. Separate the uranium from the salts using fluoride volatility or another process. Convert the UF6 to U3 O8 and place the oxide in storage in the ORNL Radiochemical Development Facility (Bldg. 3019). Interface with the uranium materials disposition program through the national 233 U repositol)' at the ORNL RDF. Remove the MSRE fuel and flush salts from their drain tanks. Stabilize the salts (by the addition of a chemical getter or by conversion to another waste forn1). Place the snits in containers and load into an
r------------, MSRE ends
-----~ Materials !
- disposition :
Salt, as-is Store in RDF Purge Convert to and trap U308 Uranium .i:- 1
°'
Melt and Fluorinate Stabilize Pack in Ship to Place in withdraw uranium with getter canister Carlsbad WIPP Salt Stop Fig. 4.3. Flowchart for alternative 4, transfer uranium to the materials disposition program and dispose of the salt residue in WIPP.
4-7 RH-TRU canister. Store the salt waste in the ORNL RH-TRU storage facilities until WIPP is available. Ship the RH-TRU canisters to Carlsbad, New Mexico, and place in WIPP for permanent disposal. The major issue is whether MSRE salt residue, without uranium, can be classified as defense transuranic waste, or if it will be classified as high-level or nondefense waste. The WIPP Land Withdrawal Act prohibits the disposal of high-level waste at WIPP. The classification of the MSRE salt is a programmatic decision. The waste acceptance criteria for WIPP are not met if the uranium is not removed from the fuel salt. They probably can be met without fluorination of the flush salt. With fluorination in place for the fuel salt, however, fluorination of the flush salt would also be feasible. 4.5 DISPOSAL OF KEY CONTAMINANTS IN SALT RESIDUE IN THE FEDERAL REPOSITORY If the salt residues are classified as high-level waste, then the federal repository, such as the repository proposed for the site under characterization at Yucca Mountain, would be the ultimate destination. This section differs from Sect. 4.3 in that it presumes that the uranium is separated from the other radionuclides in the salt. The waste acceptance criteria have not yet been defined for the federal repository. However, borosilicate glass has been extensively studied, and is being produced at Savannah River and West Valley. Borosilicate glass is as close to a standard fom1 for disposal in the repository as exists at the present time. A number of strategies can be identified for placing the key contaminants in the MSRE salts into borosilicate glass, using existing or proposed facilities. The simplest approach would be to dissolve the salts and blend them into the feed to the Savannah River DWPF. Although this would take place over an extended period of time (possibly more than 25% of the overall DWPF operating campaign), it does appear to be technically possible. The low limit on the fluoride concentration that leads to the long campaign at DWPF is likely to be common to all similar facilities. A proposal has been made to first calcine the MSRE salt solution, then vitrif)1 it at a later date. This option is identified for the case in which the uranium remains with the rest oflhe materials, and is not addressed again in this category. It could, of course, be implemented with uranium removal, but the basis for tranferring material to INEL is the spent fuel EIS; the settlement agreement between Idaho, DOE, and the U.S. Navy could prohibit the acceptance of the MSRE material without the uranium. Fluorination is considered the obvious technology for uranium removal before vitrification. A different type of technology has been proposed by Argonne National Laboratory, based on an adaptation of the electrorefiner developed for the advanced liquid metal-cooled reactor program. This technology produces a series of metal waste forms, including a metallic uranium product. These wastes are being produced at ANL-W, and activities are underway to define the disposition of the wastes from that program. Because this technology is so different than vitrification, and because the electrorefiner process itself is so readily adapted to fluoride salt processing, it is charted as a second strategy under this section. - - --------------- ,~;
4-8 Thus, two alternatives arc defined. The first, charted in Fig. 4.4, is defined as follows: Alternative 5a. Transfer the Uranium to the Materials Disposition Program, Vitrify the Salt in DWPF, and Dispose ofthe Salt Residue in the Federal Repository. Remove the MSRE fuel and flush salts from their drain tanks. Separate the uranium from the salts using fluoride volatility or another process. Convert the UF6 to ~ Qi and place the oxide in storage in the ORNL Radiochemical Development Facility. Interface with the uranium materials disposition program through the national 233U repositoiy at the ORNL RDF. Remove the MSRE fuel and flush salts from their drain tanks. Stabilize the salts with a getter for shipment. Ship the salts to Savannah River. Dissolve the salts and bleed into the waste stream being fed into the DWPF, keeping fluorine content within specification. Store the waste in glass logs at Savannah River until the federal repositoiy is available. Ship the vitrified waste to the federal repositoiy for ultimate disposal. Alternative Sb, charted in Fig. 4.5, is defined as follows: Alternative 5b. Transfer the Uranium to the Materials Disposition Program, Process the Salt by Electrorefining, and Dispose ofthe Salt Residue in the Federal Repository. Remove the MSRE fuel and flush salts from their drain tanks. Construct an electrorefiner at ORNL. Separate the zirconium and rare earths from the salt by electrorefining and convert into a metal waste forn1 to be managed along with similar wastes atANL-W. Separate the uranium and other radioactive materials from the salt. Electrorefine this material in a chloride salt electrorefiner to separate the uranium. Convert the uranium to U30 8 and place the oxide in storage in the ORNL RDF. Interface with the uranium materials disposition program through the national 233U repository at the ORNL RDF. Dispose of the chloride salt along with similar wastes generated at ANL-W. Separate the transuranics, actinides, and cesium and place in a bismuth metal waste forn1. Either qualify this waste for disposal in the federal repository, or develop a vitrification technology for this material. Separate the strontium from the salt and place in long-term storage for decay. Stabilize the salt residue and transfer to a low-level waste storage facility. 4.6 REUSE OF THE SALT Reuse of the salt is preferable to disposing of one batch of contan1inated salt and preparing another with fresh materials. However, the prospects for reuse of the MSRE salts are limited. One prospect that has been identified is a program at Los Alamos National Laboratory to develop high-power accelerator targets that can generate a significant flux of neutrons for transmutation of plutonium or actinides, or other purposes. Thus, this alternative, charted in Fig. 4.6, is defined as: Alternative 6. Transfer the Uranium to the Materials Disposition Program and Transfer the Salt to Another Program for Reuse. Remove the MSRE fuel and flush salts from their drain tanks. Separate the
~
\0 Fig. 4.4. Flowchart for alternative Sa, transfer uranium to the materials disposition program and dispose of the salt residue in the federal repository, based on vitrification at the DWPF.
Convert to .,, Store in M .-----, I
,, MSRE -----~
I atena s I I U308 RDF ends
.,. I disposition I
*------------ I Uranium Chloride Dissolve and salt Salt, as-is ,* Pack for ,,. Ship to . Mix with Electrorefine transport ANL-W ,, EBR-11 salt
. ~
Purge Zirconium -..,
*, Pack for ., Ship to .. '.:;ombine with MSRE and trap transport ANL-W ,. metal waste ends I
Uranium I I
...... Bismth I I
Melt and
~ Electrorefine*
- metal
,* Qualify or .. Store as ,,. Ship to
,------Y- ----*
I I ANL waste .,:,.. withdraw - vitrify HLW repository :managemen 0 Salt
, . Metal with strontium Dispose of as LLW
.. Store for decay Place in repository Fig. 4.5. Flowchart for alternative .Sb, transfer uranium to the materials disposition program and dispose of the salt residue in the federal repository, based on electrorefining molten fluoride salt.
MSRE I M . I I
-----~ atena s :
ends : disposition :
'------------ I
'1I Store in Salt, as-is RDF Purge Convert to and trap U308 Uranium I Melt and Fluorinate Stabilize Pack for Ship to MSRE withdraw uranium with getter shipment Los Alamos ends Salt I I I I I r------Y------,I I
- Accelerator :
- program :
*------------ I Fig. 4.6. Flowchart for alternative 6, transfer uranium to the materials disposition program and transfer the salt to another program for reuse.
4-12 uranium from the salts using fluoride volatility or another process. Convert the UF6 to U3 O8 and place the oxide in storage in the ORNL RDF. Interface with the uranium materials disposition program through the national 233U repository'at the ORNL RDF. Remove the MSRE fuel and flush salts from their drain tanks. Stabilize the salts with a getter and transport the salt to Los Alamos for use as a molten salt accelerator target. Immediate interest at Los Alamos is focused on the largely nonradioactive coolant salt. If the program develops as planned, transfer of the flush or fuel salts is possible. 4.7 INTERIM STORAGE Ultimate disposition requires the availability of repositories not yet in operation or constructed; consequently interim storage of intem1ediate or final waste fom1s will probably be required for most alternatives. Thus it is advisable to evaluate interim storage as an end point for this selection process, even though it will require further activities at a later date. Implementation of the previous strategies (aside from pemrnnent disposal in the drain tanks) requires the availability ofWIPP or the federal repositol)1* as well as the availability of processing facilities at other sites. WIPP has been the subject of a long legal and regulatOI)' struggle to bring it near to operation, and a successful completion of that process still cannot be assured. The federal repository is at a far earlier stage of the process; the Yucca Mountain site is a candidate site undergoing characterization. The INEL strategy involves a vitrification facility that is planned for construction after 2010. Although a demonstration is underway, the Argonne electrorefiner has not been tested or constructed, and linkages with waste disposal at ANL-W are in early stages. Thus, it is possible that a firm commitment to a schedule for the full strategy represented by any of these options cannot be made at this time, and a period of interim storage is required. Material placed in interim storage should be stable and pose minimal risk to personnel and the environment at the storage site. Given the uncertainties in ultimate disposition, the material should be packaged and stored in a manner acceptable for decades with minimal or no maintenance required. The storage mode should be targeted to at least one ultimate disposition strategy; flexibility in adapting to one of several strategies can be an advantage. At the same time. once the salts arc being processed, it is desirable to produce a waste fonn acceptable for ultimate disposal if at all possible. This eliminates the need for a second processing campaign at a later date. In general, the logical interim storage modes can be identified by examining the full disposition strategies, and identifying the rational hold points in each strategy. The material that exists at that hold point is then evaluated against criteria for safe interim storage. It should be noted that a strategy that leads to a desirable ultimate disposition may not exhibit an equally desirable opportunity for interim storage. There are five general modes for interim storage that appear representative. Four of these are associated with alternatives identified earlier. These are fluorination and stabilization of salt with a fluorine getter, electrorefining the MSRE salts, management as spent fuel at INEL and conversion to glass at DWPF. Some of the additional proposals for new fonns of processing at ORNL could also lead to safe interim storage; these are collected into a fifth interim storage mode. The five options that result are summarized as:
4-13 Alternative 7a, transfer uranium to the materials disposition program, stabilize the snit residue with a fluorine getter, and store at ORNL. Alternative 7b, transfer uranium to the materials disposition program, convert radioactive materials in the salt into metallic and other waste forms by electrorefining, and store at ANL-W. Alternative 7c, Calcine the salt, including the uranium, and store the calcine at INEL. Alternative 7d, transfer uranium to the materials disposition program, incorporate the salt residue in borosilicate glass, and store at Savannah River. Alternative 7e, transfer uranium to the materials disposition program, construct a salt conversion facility at ORNL, convert the salt residues to glass or phosphate waste forms, and store at ORNL. Since the first of these is the most readily implemented and the least expensive, it is used in the overall evaluation of alternatives. Alternative 7. dingramed in Fig. 4. 7, is thus identified as: Alternative 7. Transfer the Uranium to the Materials Disposition Program and Place the Salt Residue in Interim Storage. Remove the MSRE fuel and flush snits from their drain tanks. Separate the uranium from the salts by fluoride volatility. Convert the UF6 to U3O8 and place the oxide in storage in the ORNL RDF. Interface with the uranium materials disposition program through the national 233 U repositol)' at the ORNL RDF. Remove the MSRE fuel and flush salts from their drain tanks. Stabilize the salts by the addition of a chemical getter. Package the salts in a form compatible with repositol)' containers such as the RH-TRU canister. Store the salt waste in the ORNL waste storage facilities until a permanent disposition mode becomes available.
r-------------, M . I I MSRE -----~ I atena s l ends : disposition l
*------------ I Salt, as-is Store in RDF Purge Convert to and trap U308
~
Uranium I
~
Melt and Fluorinate Stabilize Pack in withdraw uranium with getter canister Salt I I I I I r------X-----,I I 1 I Future 1 I
- shipment :
'------------ I Fig. 4.7. Flowchart for alternative 7, transfer uranium to the materials disposition program and place the salt residue in interim storage.
5-1
- 5. EVALUATION OF ALTERNATIVES 5.1 EVALUATIONMATRIX Table 5.1 presents the evaluation of the most credible alternatives to each disposition end point against the CERCLA evaluation criteria and two additional programmatic criteria. A qualitative ranking system is used, scoring each alternative as high, medium, or low based on relative comparisons of the alternatives for each criterion. In some cases, specific criteria are not applicable, or the score cannot be assigned without additional information. The scores are based on the particular sequence of activities identified in the previous chapter. In some cases, notes are provided on how alternate implementations of a strategy might affect the scores.
5.2 SCORING RATIONALE FOR EACH CRITERION 5.2.1 Overall Protection of Human Health and the Environment The baseline risk presented by reactive gas currently generated in the salt and by the potential for criticality should water flood the salt in the drain tanks represents an unacceptable level of protection of human health and the environment. The no action alternative incorporates no features to mitigate the above consequences, and is thus scored as unacceptable against this criterion. In the case of enhanced storage, adequate protection is provided so long as the reactive gas removal system and the building sump pwnps remain in operation. Although this provides an adequate level of safety for interim storage, reliance on active systems cannot be assured for the lifetime of the 233 U isotope. Thus, even if the consequence of reactive gas generation can be controlled, criticality safety remains an issue. The drain tank cell can be further sealed to guard against water intrusion, and neutron poisons can be placed around the tanks or in the vapor space inside the tank above the salt. However, it cannot be assured that these measures will remain in pince for tens of thousands of years. The activities needed to integrate poisons or depleted uranium directly into the snit matrix are the same as would be required to remove the salt, and removal of the salt to a waste storage or disposal facility is a preferred activity. Thus, enhanced storage is also scored as unacceptable against this criterion. Since alternatives 1 and 2 do not provide an acceptable level of protection ofhwnan health and the environment, they are rejected outright and are not scored further against the other CERCLA criteria. Alternatives 3 through 7 are scored in terms of the level of protection provided by the form of material being stored and the final storage location. All of the options that result in the key contaminants being placed in geologic disposal rank high. It is assumed that the fissile materials disposition program will either provide for the reuse of 233U or that it will eventually be placed in geologic disposal. Likewise, it is assumed that the reuse option places tl1e salt in responsible stewardship; for the purposes of the MSRE program, this alternative is also scored as high. In alternative 7, the uranium is separated and the salt is stabilized. However, the snit remains in interim storage in a relatively accessible location, and this alternative does not provide as much protection as the repository alternatives. Therefore, alternative 7 is scored as medium. 5.2.2 Compliance with Applicable or Relevant and Appropriate Requirements It is assumed that all of the alternatives are executed in a way that complies with all applicable or relevant and appropriate requirements. Thus, all of the options are scored as high.
5-2 Table 5.1. Alternatives evaluation matrix for the Molten Salt Reactor Experiment Fuel Salt Disposition task Altcrnntive0 CERCLA evaluation criteria 2 3 4 Sa Sb 6 7 Overall protection of human health and the u u H H H H H M environment Compliance with applicable or relevant and X X H H H H H H appropriate requirements Long-term effectiveness and pennanence X X H H H M na L Reduction of toxicity, mobility, or volume through X X H M H H na M treatment Short-term effectiveness X X H H H H H H Implementability X X L L L L ? H Cost (scored as affordability) X X L M L L ? H Other evaluation criteria: Adaptability to changes in system requirements X X H H H M H H Compatibility with programmatic objectives L L H H H H H M "Key to alternatives:
- 1. No action.
- 2. Enhanced storage.
- 3. Dispose of the salt. including uranium, in the federal repository.
- 4. Transfer the uranium to the materials disposition program and dispose of the salt residue in WIPP.
Sa. Transfer the uranium to the materials disposition program and dispose of the salt residue in the federal repository (via DWPF). Sb. Transfer the uranium to the materials disposition program and dispose of the salt residue in the federal repository (via electrorelining).
- 6. Transfer the uranium to the materials disposition program and transfer the salt to another program for reuse.
- 7. Transfer the uranium to the nmterials disposition program and place the sail residue in interim storage.
Key to matrix scores: L Low M Medium H High U Unacceptable na Not applicable x Not scored ? Unknown However, regulatory and policy issues may affect the implementability of the various alternatives. In particular, compliance with the WIPP Land Withdrawal Act may preclude classification of the salt residues as RH-TRU waste. These considerations are reflected in the scores assigned to implementability. 5.2.3 Long-Term Effectiveness and Permanence High scores are assigned to those alternatives that place the key contaminants in ultimate geologic disposal or place the uranium under the management of the fissile materials disposition program as an oxide. A low score is assigned to the interim storage alternative. The reuse alternative is not scored against this criterion.
5-3 5.2.4 Reduction of Toxicity, Mobility, or Volume Through Treatment A high score is given to alternative 5a, which results in a well-qualified borosilicate waste form. Although the exact waste form for alternative 3 has not been defined or tested, it too is assumed to be borosilicate glass and is scored high. The waste forms produced by alternative Sb vary, and have not been qualified. Until they are qualified, it is not clear that they represent a final end point. However, because the volwne is low, alternative 5b is scored high. Scores of mediwn are assigned to alternatives that do separate uraniwn from the salt and that have stabilized salt as an endpoint (alternatives 4 and 7). The reuse alternative is not scored against this criterion. 5.2.5 Short-Term Effectiveness Alternatives 4 through 7 all remove the salt, separate uraniwn from the salt, place the uraniwn in storage as an oxide, and as a minimum place stabilized salt in storage in a diy location. All of these alternatives prevent further migration of uranium, and control any additional radiolytic fluorine. These alternatives are scored as high. Alternative 3 does not remove the uraniwn from the salt, but preswnably converts it to an oxide in the calcination step, and immobilizes it in glass in the final immobilization step. Thus, this alternative is also scored as high. 5.2.6 Implementability Alternatives 3, Sa, and Sb all rely on the federal repository for spent fuel and high-level waste, which is years away from construction. Alternative 4 relies on WIPP, which although constructed is not yet receiving waste. The classification of salt residues as RH-TRU may not be possible, in light of the WIPP Land Withdrawal Act and associated guidance on distinguishing RH-TRU from HLW. Alternative 3 relies on the future funding and construction of an immobilization (preswnably vitrification) facility at INEL. Dissolving and blending the salts directly"into the feed to the DWPF requires such a long, gradual bleed that if any difficulties are encountered (such as existing material in the Savannah River waste that further reduces the acceptable feed rate) it may fail on technical grounds. The electrorefining technology itself can probably be implemented, but the trail of activities needed to fully dispose of the wastes produced by the initial electrorefining step is not fully developed. Thus, all four of these alternatives are scored as low in terms of implementability. Four of the alternatives are ranked low in implementability for different reasons. Consequently, it is difficult to rank the four comparatively. WIPP is further along than the federal repository, and the impediments to implementation of the WIPP option are legal and programmatic. Thus, alternative 4 might stand a better chance than blending directly into an existing borosilicate glass facility. Although the initial calcination step associated with alternative 3 may prove credible (at least for the fuel salt, with transfer to Idaho made under the spent fuel EIS), further processing into a final waste form would be in a facility not yet funded and disposal in a repository not yet sited. Similar issues apply to alternative Sb. Without transfers of waste to Idaho for storage, the alternative cannot be implemented as defined. The settlement agreement between Idaho, DOE, and the Navy may make such transfers difficult. Alternative 6 relies on a program that can make use of the salts. Programs such as the accelerator-driven transmutation project at Los Alamos are in the planning stage and are not yet funded. The probability of such programs accepting salt is a factor of the maturity of the program and the salt in question. Use of the nonradioactive coolant salt may be relatively likely. Use of the highly radioactive fuel salt may in general be relatively unlikely, although san1ples may be of use by researchers developing salt cleanup systems. At this time, implementability of alternative 6 is left as a question mark.
5-4 Implementability of the interim storage option is, of course, higher. Alternative 7, based on interim storage on the ORNL site, is scored as high. All of the technologies needed to implement this alternative have been practiced at some time or other in the histol)' of the MSRE project, or are common uranium processing technologies. Only the salt chemist!)' issue poses a technical implementability question, and activities are already underway to bound the potential consequences of this issue and to ensure that appropriate technology is available to safely melt the salt. The main implementability issue is associated with the acceptability of interim storage as an option. It is crucial that this storage option be defined in the context of a credible ultimate disposal strategy. 5.2.7 Cost This criterion is scored as affordability, so that high is a desirable ranking as in the case of all other criteria. The relative cost of the alternatives can be evaluated by comparing the number of activities needed to implement the alternatives. In each of alternatives 3 through 7. removal of the salt is necessal)'. The cost of salt removal is assumed to be an equal baseline cost for each alternative, and removal cost is not a discriminator. For all scenarios except alternatives 3 and 5b, the uranium is separated by fluorination and converted to oxide. Because fluorination is a simple process (as seen by the relatively small equipment in the present fuel processing cell), and because an oxide conversion facility is already being constructed for the project, this cost is seen as being relatively moderate. The cost of gettering is assumed to be insignificant. Design of an external getter may be no more difficult than including a chemical filter in series with a HEPA filter in a package vent. If a getter material is integrated into the salt matrix, the cost of gettering (development aside) would likely only be the cost of adding the material to the salt before it is allowed to freeze. For alternative 7, the only other cost incurred is storage. In the vel)' long haul, storage costs will accrue. But for any near-term cost estimate, the affordability of alternative 7 is high. For alternative 4, the only other cost incurred is shipment to WIPP. This entails transport in a licensed cask from Oak Ridge to Carlsbad, New Mexico. Because transportation issues can be costly to resolve, this alternative is scored as medium in terms of affordability. If implementation can be based on an existing, licensed WlPP cask and all transportation issues are resolved by the transuranic waste program, the affordability of this alternative might be raised to high. In the case of alternative 3, fluorination and conversion is not required. However, a fissile material must be shipped from Oak Ridge to Idaho. No existing program is likely to pay the costs for cask licensing or other transportation issues. Thus, the affordability of this alternative would be no higher than medium. But on top of these issues, a dissolver and feed system must be constructed to interface with the calciner facility. This dissolver would need to be shielded and provided with an appropriate off-gas system. An existing facility at the ICPP might be used, but this would incur the cost of restarting an ICPP facility. A new facility could also be constructed for this purpose. The real cost does not end with calcining, although the waste becomes indistinguishable from other INEL waste once it is mixed and sprayed into the calciner. The affordability of this alternative is ranked low.
5-5 In a similar way, the alternative lo utilize the DWPF begins with all of the activities associated with storing and shipping fluorinated, stabilized salt, and then adds the cost of a dissolver facility and an extended operating campaign. This alternative is also ranked as low in terms of affordability. The cost of constructing and operating an electrorefiner cover the construction of an inert glovebox, shielding around the glovebox, an appropriate off-gas system, further refining of the uranium product, and disposal of the electrorefiner wastes. Thus, alternative 5b is ranked as low. Until the specific activities associated with a reuse option are identified, a score is not given for the cost of alternative 6. If implemented, the cost is likely to be relatively low. 5.2.8 Adaptability to Changing Requirements Because no repository is currently operating and acceptance criteria are still being developed, the ability to adapt to changes in waste forn1 and handling requirements is a valuable attribute. At present, any processing that leads to an alternate waste fonn begins with salt. Thus, any alternative that ends with stabilized salt likely preserves all processing alternatives that are presently available. Uranium oxide is a common form for both storage and as a feed for other processes. Uranium fluoride is also a common feed to other processes. Thus, alternatives that leave the uranium as oxide or as a fluoride salt preserve a wide range of processing alternatives. Alternatives 4, 7, and probably 6 all result in processing end points consistent with the above, and are scored as high. Alternatives 3 and 5a result in the salt residue being placed in a borosilicate glass waste form. It is very difficult to remove the materials from this waste form and convert them to another; thus these alternatives might be scored as low. However, it is also very unlikely that the borosilicate glass waste form would not be accepted under any future waste forn1 acceptance criteria. Thus, alternatives 3 and Sa are scored as medium. Alternative Sb results in a variety of metal waste forms. Further processing flowsheets for these waste forms are limited at this time. The waste forms are not yet qualified for disposal. Thus, this alternative is scored as low. (On the other hand, the technology itself could be modified to fit within other waste management strategies. Since this does not fit within the definition of the baseline strategy, it is not a basis for scoring.) 5.2.9 Compatibility with Programmatic Objectives Programmatic objectives include meeting the DNFSB goals of removing the salt from the drain tanks and placing the uranium in storage as oxide, and allowing the decommissioning of the MSRE facility to proceed. Achieving an ultinmte disposition of the salt, with no further cost incurred for interim storage, is also a programmatic objective. The repository and reuse alternatives accomplish all these programmatic objectives, and are scored as high. Alternative 7 meets all objectives except eliminating further storage cost for the salt, and is scored as medium. Alternatives I through 3 do not result in the separation of uranium from the salt and conversion to oxide, as specified in the DNFSB 94-1 Implementation Plan. Alternative 3 does, presumably, convert the uranium to oxide without separation. It is thus given the same score as alternatives that convert uraniwn to oxide after separation. Alternatives I and 2 do not result in removal of the salt from the drain tanks, as specified in the DNFSB 94-1 Implementation Plan, and as needed to continue with facility decommissioning. These alternatives do not eliminate further storage costs for the ~alt. Thus, alternatives I and 2 are scored as low.
5-6 5.3 RECOMMENDATIONS The reuse alternative (alternative 6) and the geologic disposal alternatives (alternatives 3, 4, Sa and Sb) score the highest, but all have significant implementability issues. Reuse of the MSRE salt is preferable to disposal ofMSRE salt and production and contamination of a new batch of salt. In the case of reuse at Los Alamos, transfer of the salts is contingent upon funding for the accelerator-driven transmutation program. Transfer of the nonradioactive coolant salt may occur upon funding for the research phase of the project, and is relatively probable. Samples of fuel or flush salt may also be wanted in the research phase to test salt cleanup systems. Full transfer of the fuel and flush salts involves a more significant investment in storage and cleanup at Los Alamos; availability of such funding is likely contingent on approval of a line-item to design and construct the accelerator. The relative costs of cleaning up and reusing the MSRE salt (if clean salt is required) as opposed to preparing a new batch of salt will also be a factor. In evaluating this tradeoff, the cost savings of not disposing of MSRE salts should be taken into consideration. Alternatives 4, Sa, and 5b transfer uranium to the fissile materials disposition program in a generally accepted forn1, and provide for true disposal of the salt residues. Alternative 3 does not separate the uranium, but is presumed to feed it into a vitrification facility as an oxide. Alternatives Sa and Sb, based on disposal in the federal high-level waste repository, cannot be fully implemented until the repository is constructed. Incorporation into the DWPF feed and storage in borosilicate glass at Savannah River might be implemented in this time, resulting in interim storage of a waste form clearly targeted for the repository; this is one of the alternative versions of the interim storage option. Similarly, alternative 3 could produce interim storage first of calcine, later of glass. Disposal at WIPP is more credible in the near future as WIPP has been constructed and is scheduled to begin receiving RH-TRU waste early in the next decade. However, operation ofWIPP is still being challenged and acceptance criteria may change as EPA repository perfonnance guidance is interpreted. Furthermore, although the salt residues appear to meet WIPP acceptance criteria by characteristic,the classification of this unique waste form with respect to the definitions in the WIPP Land Withdrawal Act is unclear. Even in the strategy defined here, alternative 5b does not fully define the path to ultimate disposal for all of the waste forms produced. Thus, none of the alternatives 3. 4, 5a. 5b, and 6 can be fully implemented at this time. Alternatives I and 2 (no action and enhanced storage) do not provide adequate protection of human health and the environment for final disposition, and are thus rejected. Reasonable measures can be implemented to control reactive gases and prevent criticality in the near term, providing safe storage until the disposition activities are implemented. However, enhanced storage does not meet programmatic objectives, and as equipment ages, removal of the salts from the tanks becomes more difficult. Interim storage of salt residues after the uranium is separated and converted to oxide can be implemented in a number of ways. The result is storage of salt residues in a monitored facility, but the storage mode can be tailored to the most likely ultimate disposition path. Uranium is separated, converted to oxide, and transferred to the fissile materials disposition program. This addresses the most serious near term concern (that of 232 U and its daughters), and if ultimate disposal by the materials disposition program is assumed, it also addresses the most serious very-long-term problem (the radionuclide inventOI)' associated with the 233 U decay chain). The base path chosen for the alternatives evaluation is storage of fluorinated and stabilized salt residues at ORNL, in a form compatible with ultimate disposition in WIPP. Other versions of this alternative are considered in the next section. The waste form is sufficiently stable that a reasonable level of protection of human health and the
5-7 environment is achieved. Generation ofUF6 is no longer possible, and generation of gaseous fluorine is chemically controlled. All programmatic objectives except the elimination of future storage and monitoring costs are met. This alternative can be implemented now with a modest investment in facilities at the MSRE and at the site chosen for conversion of other MSRE uranium. In the end, all of the options except those that leave the salt in the drain tanks are reasonable. Interim storage does not complete the job, and should be selected only as a temporary measure. The most significant discriminator for all other alternatives is implementability. The following sequence of preferences can be defined: If reuse of the fuel and/or flush salts is a serious option, then it should be selected. If any of the geologic disposal alternatives prove technically and programmatically implementable, every reasonable effort should be made to implement that alternative. In particular, efforts should be made to resolve the regulatory and programmatic obstacles to disposal in WIPP. At the present time, only the interim storage option is likely to be implementable. It does provide adequate protection of human health and environment, but at some time an ultimate disposition must be identified to end the cost of perpetual care and monitoring. Storage of fluorinated, stabilized salt can meet the quantitative requirements for disposal at WIPP, or can serve as the feed for any of the other processes evaluated. Thus, interim storage is recommended as being consistent with ultimate disposition and being fully implementable at this time. 5.4 REVIEW OF THE ALTERNATE APPROACHES TO INTERIM STORAGE Because the previous evaluations were performed in the context of a series of activities that, in most cases, end with geologic disposal, the application of a subset of those actions to provide an interim storage mode was not formally evaluated in the previous matrix. Since that matrix identified interim storage as the only credible alternative for implementation at this time, a brief review of a set of alternatives for interim storage is presented here. Table 5.2 presents an evaluation matrix for interim storage alternatives. These alternatives were defined in Sect. 4.7. Since all of these strategies are for interim storage, many of the CERCLA scores are nearly identical, and the overall scoring tends to be compressed. Because the scores are relative to the other interim storage options, alternative 7a does not necessarily score the same as in the main matrix. Three additional criteria were added to the interim storage evaluation matrix. One evaluates the stability of the material that is placed in interim storage. The second addresses the ease (and thus the long-term cost) with which the material can be stored. The last addresses the adaptability to changes in the ultimate disposition strategy. All of the options result in stable material in interim storage, and are scored as medium against the overall protection of human health and the environment criterion. Similarly, all will be performed in a manner that fully complies with all applicable, relevant, or appropriate requirements, and are scored high. None of these options provides for ultimate disposition of the key contaminants, and thus none is scored as high against long-tern1 effectiveness and permanence. The DWPF glass is likely to be
5-8 acceptable at any disposal facility, and is thus scored as medium. The immediate products of the electrorefiner and the INEL calcine both require further processing before disposition, and are scored low. If the fluorinated, stabilized salt can be disposed ofWIPP, the interim storage option is not needed. Ifnot, the stored material will probably need further processing to fit into any other strategy. Thus, the long-term effectiveness and pennanence for alternative 7a is also scored as low. Final definition of the waste forms from a new process cannot be evaluated at this time. The reduction of toxicity, mobility, or volume through treatment is reasonable for any alternative. Thus, each of the first four alternatives is scored at least medium. Only the borosilicate glass form generated by DWPF has been demonstrated in a rigorous qualification program, and thus alternativ e 7d is scored as high. Because of the low volwne of wastes generated by alternative 7b, it too is scored high. Again, the waste forms from the new processes have not yet been evaluated. All of the alternatives prevent recurrence of the difficulties that have led to the MSRE Remediation Project. Thus, each is scored high in terms of short-term effectiveness. The new processes represented by alternative 7e have not yet been developed, and their implementability is scored as low at this time. Moderate investments in studies of these technolog ies might result in higher scores. Alternative 7a requires only modest activities at the ORNL site, and no significant impediments to implementation are known. Alternative 7b requires the construction of a (moderate) electrorefiner facility and integration with waste management at another site. and alternativ es 7c and 7d involve transportation of salt to another state. The implementability of these is ranked as medium. Technical challenges might drive the implementability score for DWPF to low. The electrorefiner alternative incurs the cost of constructing a moderate processing facility, and possibly the transportation of wastes to ANL-W for common management with wastes generated by processing EBR-11 fuel. The affordability of alternative 7b is presently scored as low. Implementation of calcining at INEL or vitrification at Savannah River incur both the costs of transportation to another site and construction of a dissolution and feed facility, and are also scored as low. Alternative 7a incurs the costs of fluorination, using an existing conversion facility to convert to oxide, and storage with a getter. None of these requires a major new facility, and no transportation costs are incurred. The affordability of alternative 7a is ranked as high. Insufficient descriptions of the facilities required to implement the GM ODS or phosphate processes exist to score alternative 7e. The salt can be stabilized with a getter; the package is stable but since radiolysis continues alternative 7a is scored as medium against stability of. waste form. Metal and salt wastes from electrorefining also*may need further stabilization, and that waste form is scored as medium. Calcine is presumed to be stable but is soluble and granular; it too is scored as medium. Glass is very stable; alternative 7d is scored as high. Insufficient data exists to score the stability of the new approach es. Although the salt package defined in this report would appear to be relatively straightforward, such storage is not trucing place at ORNL today (MSRE drain tanks excepted). Ease of storage and handling is scored as medium. If the RH-TRU handling facilities at ORNL can be used for the salt, and adapt to the standard RH-TRU configuration, this score might be raised. Likewise, the electrorefiner process produces a host of unusual wastes (none of which, individually, should be difficult to handle) not presently managed at ORNL, and this alternative is scored as medium. The INEL calciner and DWPF glass waste forms are indistinguishable from other wastes generated at those facilities, and both calcine bins and glass storage facilities exist. These are scored as high. Again, handling issues with the novel waste forms are difficult to score at this time. However, no clear difficulties with either GM ODS product or the phosphate waste forn1s are envisioned.
5-9 Table 5.2. EYaluation matrix for alternate approaches to alternati\'e 7, transfer uranium to the materials distribution program :me.I place the salt residues in interim storage Alternativea CERCLA evaluation crite1*i11 7a 7b 7c 7d 7e Overall protection ofhumnn health and the environment M M M M M Compliance with applicable or relevant and appropriate requirements H H H H H Long-term effectiveness and pemianenee L L L M ? Reduction of toxicity, mobility, or volume through treatment M H M H ? Short-term effectiveness H H H H H Implementability H M M M ? Cost (scored as affordability) H L L L ? Other evaluation criteria: Stability of the waste form M M M H ? Ease of storage and handling M M H H ? Adaptability to altemate ultimate disposition strategies H M L H H
?
- Key to alternatives:
7n. Transfer uranium to the materials disposition progmm. stabilize the snit residue with a getter. and store at ORNL 7b. Transfer uranium to the materials disposition progmm. convert radioactive materials in the snit into metallic and ceramic wru.1e forms by electrorefining. and store at ANL-W. 7c. Calcine the fuel salt, with uranium, and store at INEL 7d. Transfer uranium to the materials disposition program. incorporate the salt residue in borosilicate glass, and store at Savannah River. 7e. Transfer uranium to the materials disposition program, construct a snit conv.:rsion facility at ORNL convert the snit residues to glass or phosphate waste fonns. and store at ORNL Key to matrix scores: L Low M Medium H High ? Unknown Because interim storage is only selected if a path to ultimate disposition cannot be achieved, it is an obvious advantage for the material stored to be adaptable to several different ultimate disposition strategies. Fluorinated salt is the beginning of most of the processing scenarios, and thus this form is scored high against adaptability. The electrorefiner produces metals and a salt; these waste forms are not unusual materials but they do not appear in most of the ultimate disposition flowsheets. A score of medium is assigned to the electrorefiner wastes. Calcine blended with the INEL waste stream is compatible only with the INEL waste disposition strategy; its adaptability is scored as low. Similarly, the DWPF glass option dedicates the MSRE material to the Savannah River strategy, but since borosilicate glass is considered acceptable at most sites it is scored as high. The GMODS process could produce glass product, or could produce feed material for an existing vitrification process. The phosphate process produces a unique product that may or may not be adaptable to other processing scenarios. Thus, an overall score of high, with an indication of uncertainty, is given to the new technologies. In the end, implementability and affordability are the major factors that result in a recommendation for alternative 7a. Alternative 7d offers a better, borosilicate glass waste form, but its implementability
5-10 score is not high, and its affordability (with a long feed campaign) is likely to be low. The calcine waste fonn is mainly attractive in tenns of creating a waste that is already being manage d by the DOE system. Electrorefining creates a variety of relatively simple but unique waste fom1s, and incurs the cost of constructing a new facility. The novel flowsheets (GMODS or phosphate conversion) might offer advantages over storage of fluorinated, stabilized salt, but further development be demonstrated. is needed before this can
6-1
- 6. IDENTIFICATION OF KEY UNKNOWNS, POTENTIAL CONSEQUENCES, AND PATH TO RESOLUTION 6.1 CERCLA PROCESS This report is being issued to aid in the preparation of a CERCLA feasibility study for an interim remedial action. The CERCLA feasibility study *will be followed by a proposed plan, and will eventually lend to a record of decision. It is the CERCLA record of decision that formally commits to a given alternative. It is currently expected that the record of decision will be issued in May, 1998. This report only provides recommendations for use in the CERCLA process.
The CERCLA process therefore lends to an uncertainty in which alternative will actually be selected. It is possible that as the CERCLA process continues, with input from regulators and the public, that new perspectives or even new alternatives will be identified. Therefore, the final record of decision may not reflect the recommendations of this report. 6.2 WASTE CLASSIFICATION The overall logic of this evaluation is to define the desired end point first, and then to identify credible strategies for arriving at that end point. The transuranic content of the salts indicates that geologic disposal is appropriate, at lenst for the transuranic isotopes. The WIPP has been constructed and intends to begin operation with contnct-hnndlcd waste in 1998, nnd with remote-handled waste around 2002. The federal repository, designated for the disposal of spent fuel and high-level waste, is still in the site characterization phnse and will not open until after 20 I 0. Thus, WIPP is an attractive end point for the key contaminants in the snit. WIPP has been constructed for disposal of defense transuranic waste. The WIPP Land Withdrawal Act prohibits the disposal of high-level wnste at WIPP. Fluorinated MSRE salt residues, or other waste forms that can be produced, can meet the quantitative requirements that are presently identified for disposal of RH-TRU at WIPP. If the MSRE waste is classified as spent fuel or high-level waste, however, such disposal would not be allowed. The fission product concentration falls below the quantitative activity concentration limit in the WIPP waste acceptance criteria, and the fission product concentration nearly falls into the range that could be classified as Class C waste under IO CFR 61 (although the transuranic content greatly exceeds the.criteria for near-surface disposal). The total amount of waste is small (8 RH-TRU canisters each of fuel and flush salt residue are projected) and the total radionuclide inventory in the disposal site after 500 years is about 80 Ci. The overall impact on a facility such as WIPP is minimal. If the MSRE waste can be classified based on its characteristics, a transuranic classification appears reasonable. If classified on the basis of history, a more complex processing strategy and a long period of interim storage results. A similar issue is associated with the definition of "defense-related." Similar waste classification nnd acceptance issues would be faced by waste forms other than salt from which the uranium has been removed, such as the waste forms produced by the electrorefining technique.
6-2 6.3 SALT CHEMISTRY An improved W1derstanding of several salt chemistry issues is needed to support the methodology for melting the salt, especially the fuel salt. An improved understanding of the performance of getter materials could lead to better strategies for controlling the further production of radiolytic fluorine, including strategies that integrate a getter into the salt matrix to prevent any evolution of fluorine without external controls. A plan for a series of experiments is being developed to verify the existing knowledge and assumptions of salt behavior. These experiments will
- Establish the significance (or insignificance) of melting highly reduced salt,
- Evaluate the potential for low temperature annealing treatments with F to re-establish salt 2
chemistry,
- Demonstrate an assured gettering strategy,
- Establish confidence in the pool-melt strategy, and
- Continue to strive to identify the mechanism ofUF6 fonuation.
In addition, a priority will be given in the next fiscal year to activities that will access the drain tank cell, inspect equipment in the cell, and (once the reactive gases have been removed) access a drain tank for inspection of the salt mass. This *will provide additional infomiation on the present physical condition of the salt, and will help to assess the condition and operability of equipment that could be used to remove the salt. 6.4 AVAILABILITY OF SUPPORTING FACILITIES Several of the alternatives presented in this report utilize existing facilities or sites for the storage, processing and disposal of the radionuclide-inventory in the salts. The availability and appropriateness of using these facilities would affect the implementation of those strategies. The availability of WIPP or the federal repository is obviously a key issue, both in tem1s of the facility entering operation and in terms ofMSRE waste being allowed into the disposal facility. Strategies that make use of existing facilities such as the DWPF or the INEL calciner and eventual waste immobilization facility depend on the availability of such facilities. This includes agreements with the sponsor that the MSRE materials can be processed in that facility, and verification that the use is technically acceptable. In some strategies, construction of new facilities (such as the INEL immobilization facility) are assumed: implementation of that strategy is thus contingent on the fW1ding and construction of that facility. 6.5 TRANSPORTATION AND INTERSTATE AGREEMENTS Transportation of radioactive waste to another state for processing, storage, or disposal is a sensitive issue. States' equity issues may have to be resolved between a host state and the State of Tennessee, or between a host state and the federal government. In some cases, these relationships are highly litigated and controlled by court orders or settlement agreements (such as the Idaho settlement agreement). With other states, these agreements arc arrived at on a case-by-case basis. Special purpose casks may be required, involving licensing agencies.
6-3 6.6 FUNDING Implementation of any strategy is contingent upon the availability of funding for the necessary activities. This includes funding for activities perfom1cd within the scope of the fuel salt disposition task, as well as funding for key interfacing facilities such as proposed processing or disposal sites.
7-1
- 7. REFERENCES Advanced Integrated Management Services, Inc. 1995. Conversion oJMolten Salt Reactor Experiment 233 UF6 for Permanent Storage, prepared for Martin Marietta Energy Systems, Inc., Oak Ridge Natl. Lab.
Advanced Integrated Management Services, Inc. 1996. Process Support and Analysis for Fuel Salt Disposition, prepared for Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab. Anon. 1995. Technical Bases for Yucca Mountain Standards, National Academy Press. Baes, C. F. 1969. "The Chemistry and Thermodynamics of Molten Salt Reactor Fuels," Nucl. Metal. 15, 617-44. Baxter, R. G. 1983. Description of Defense Waste Processing Facility Reference Waste Form and Canister, DP-1606, Rev. I. Blumberg, R., and E. C. Hise 1968. MSRE Design and Operations Repon. Part X-Maintenance Equipment and Procedures, ORNL/TM-910, Union Carbide Corp., Oak Ridge Natl. Lab. Briggs, R B., et al. 1964. Molten-Salt Reactor Program Semiannual Progress Report for the Period Ending July 31, 1964, ORNL-3708, Union Carbide Corp., Oak Ridge Nall. Lab., pp. 252-87. Brown, K. G. and R L. Postles N.d. SME Acceptability Determination for DWPF Process Control (U), WSRC-TR-95-0364, Rev. 3. Brown, J.C. July 11, 1996. Programmatic Director of LANCE and Energy Research Programs at Los Alamos National Laboratory, letter to L. M. Toth, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab. Cagle, C. D., and L. P Pugh 1977. Decommissioning Study for the ORNL Molten Salt Reactor Experiment, ORNL/CF-77/391, Union Carbide Corp., Oak Ridge Natl. Lab. Committee on Eletrometallurgical Techniques for DOE Spent Fuel Treatment 1995. An Assessment of Continued R&D into an Electrometallurgical Approach for Treating DOE Spent Nuclear Fuel, National Research Council, Washington, D.C. Crume, C. C. October 11, 1994. "Nuclear Criticality Safety Issues Associated with Uranium Migration in the MSRE," presentation to the Independent Re\'iew Panel. Denney, R D. July 23, 1996. Transmittal of Phase I Report for Memorandum Purchase Order# 10X-KEPO6V-RDD-l 2-96," letter to F. J. Peretz, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab. Denney, R D., N.d. "Transmittal of Phase II Report for Memorandum Purchase Order# 10X-KEPO6V-RDD-12-96," letter to be issued to F. J. Peretz, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab.
7-2 DeVore, J. R. July 17, 1996a. Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab.,
"Calculation of MSRE Fuel Snit Activity in a Potential Repository," letter to F. J. Peretz, Oak Ridge Natl. Lab.
DeVore, J. R. July 31, 1996b. Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab.,
"Calculation of Radiation Dose Rates from MSRE Fuel and Flush Salt Activity in a Potential Repository Package," letter to F. J. Peretz, Oak Ridge Natl. Lab.
DeVore, J. R. N.d. Description of Potential Fluoride Volatility Processing of the MSRE Fuel and Flush Salts in the Fuel Processing Cell at M.S1lE, XOE-784, August I 996 draft. DeVore, J. R., et al. 1996. Fuel Processing Cell Uranium Migration Investigation, XOE-778. DOE (U.S. Department of Energy) 1995. Department ofEnergy Programmatic Spent Nuclear Fuel Management and Idaho National Engineering Laboratory Environmental Restoration and Waste Management Programs Final Environmental Impact Statement, DOE/EIS-0203-F. DOE (U.S. Department of Energy) Carlsbad Arca Office 1995. Strategic Plan, DOE/WIPP 93-025, Rev.I. DOE/CAO 95-1095 1995. Remote-Handled Transuranic Waste Study. DOE/WIPP-069 1996. Waste Acceptance Criteria for the Waste Isolation Pilot Plant, Rev. 5. Forsberg, C. W., and E. C. Beahm 1996. Use ofthe Glass Material Oxidation and Dissolution System (GMODS) for Conversion ofMolten Salt Reactor Experiment Salts to Borosilicate Glass. Guymon, R.H. 1971. MSRE Procedures for the Period Between Examination and Ultimate Disposal (Phase Ill ofthe Decommissioning Program), ORNUfM-3253, Union Carbide Corp., Oak Ridge Natl. Lab. Hightower, J. R., et al. N.d. Low-Pressure Distillation ofa Portion ofthe Fuel Carrier Salt from the MSRE, ORNL-4577, Oak Ridge Natl. Lab. Jensen, R. J. March 2, 1995. Letter to L. M. Toth, Martin Marietta Energy Systems, Inc., Oak Ridge Natl. Lab., transmitting enclosed infommtion on Los Alamos rock melters. Kocher, D. C., N.d. Legal and Regulatory Requirements for Disposal of 233 U, ORNL/MD-45, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab., draft. Kring, C. T., and D. J. Richards N.d .. Evaluation ofMechanical Alternatives for the Removal ofthe Molten Salt Reactor Experiment Fuel, in preparation. Laidler, J. J. May 24, l 996a. Letter to D. Smith transmitting descriptions of the Argonne electrochemical process for treatment of MSRE fuel and flush salts. Laidler, J. J. June I 0, 1996b. Letter to D. Smith transmitting a cost estimate for implementing the Argonne electrochemical process for treatment of MSRE fuel and flush salts.
7-3 Lindauer, RB. 1969. Processing ofthe M.S1lli Flush and Fuel Salts, ORNL/TM-2578, Union Carbide Corp., Oak Ridge Natl. Lab. Lingle, W. N. June 12, 1995. '*Oak Ridge National Laboratory Molten Salt Reactor Experiment Facility Time Critical Removal Action Memorandum Report," letter to V. Weeks and D. McCoy, Martin Marietta Energy Systems, Inc., Oak Ridge, Tenn. Nix, C. E. June 4, 1992a. "Resource Conservation and Recovery Act Compliance Assessment of Molten Salt Reactor Experiment," letter to T. W. Burwinkle, Martin Marietta Energy Systems, Inc., Oak Ridge, Tenn. Nix, C. E. September 9, 1992b. "Resource Conservation and Recovery Act Compliance Assessment of Molten Salt Reactor Experiment-Additional Comments," letter to T. W. Burwinkle, Martin Marietta Energy Systems, Inc., Oak Ridge, Tenn. Nordhaus June 1996. '*Interpretation of the Tenn 'Atomic Energy Defense Activities' as Used in the Waste Isolation Pilot Plant Land Withdrawal Act." memorandum to G. Dials. Notz, K. J. 1985. Extended Storage-in-Place ofMSRE Fuel and Flush Salt, ORNLffM-9756, Martin Marietta Energy Systems, Inc., Oak Ridge Natl. Lab. Pub. L. 96-164, 93 Stat. 1259. U.S. Department of Energy National Security and Military Applications of Nuclear Energy Authorization Act of 1980. Pub. L. 102-579, 106 Stat. 4777. WIPP Land Withdrawal Act. Reis, V. H., N.d. '*Use of Building 3019A for Storage of Uranium Recovered from the Molten Salt Reactor Experiment," letter to Lehman. Robertson, R C. 1965. M.Wlli Design and Operations Report, Part I, Description ofReactor Design, ORNLffM-728, Union Carbide Corp., Oak Ridge Natl. Lab. Rosenthal, M. W., et al. 1971. Molten-Salt Reactor Program Semiannual Progress Reporl for the Period Ending Febmary 28, 1971, ORNL-4676, Union Carbide Corp., Oak Ridge Natl. Lab. Settlement agreement N.d. Agreement between.the State of Idaho, through the Attorney General, and Governor Phillip E. Ball in his official capacity; the Department of Energy, through the General Counsel and Assistant Secretary for Environmental Management; and the Department of the Navy, through the General Counsel and Director, Naval Nuclear Propulsion Program, entered into on October 16, 1996 to fully resolve all issues in the actions Public Service Co. of Colorado v, Batt, No. CV-91-0035-S-EJL (D. Id.) and United States v. Batt, No. CV-91-0054-S-EJL (D. Id.) Shaffer, J. H. 1971. Preparation and Handling of Salt Mixtures for lhe Mo/Jen Salt Reactor Experiment, ORNL-4616, Union Carbide Corp., Oak Ridge Natl. Lab. Sherrill, G. W. December 1, 1995. '*Preliminary Layout for MSRE UF6 Conversion Process," letter to R L. Faulkner. Shor, J. T. 1996. Compendium of Recent Building 7503 Data, ORNL/M-4808, Lockheed Martin Energy Systems. Inc., Oak Ridge Natl. Lab.
7-4 Skipper, D. D. June 4, 1992. Martin Marietta Energy Systems, Inc., Oak Ridge Natl. Lab.,
"EBASCO Compliance Assessment of the Molten Salt Reactor Experiment," letter to file.
SSA/7503-ERP/003/R0 1996. System Safety Analysis, Molten Salt Experiment Facility Interim Vent and Trap System. Thoma, RE. 1971. Chemical Aspects oJM..<:,RE Operations, ORNL-4658, Union Carbide Corp., Oak Ridge Natl. Lab. Toth, L. M. 1996. "ORNL Foreign Trip Report," ORNL/FTR-5817, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab. Toth, L. M., and L. K. Felker 1990. "Fluorine Generation by Gamma Radiolysis of a Fluoride Salt Mixture," pp. 20 l-10 in Radiation Effects and Defects in Solids, Vol. 112. Toth, L. M., and D. F. Williams 1996. POOIMELT: Chemically Controlled Melting ofMSRE Fuel Salt to Reestablish Homogeneity ofthe Melted Fuel for Safe Removal, ORNL/CF-96/37, Lockhee d Martin Energy Systems, Inc., Oak Ridge Natl. Lab. Watkins, D. Januaiy 29, 1996. Manager of the National TRU Program, "Recom mendations for Distinguishing Remote Handled Transuranic from High Level Waste," (draft policy recommendations), letter. Williams, D. F., et al. 1996. A Descriptive Model of the Molten Salt Reactor Experim ent After Shutdown: Review ofFY 1995 Progress, ORNL/TM-13142, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab. Williams, D. F., et al. N.d. Chemical Interactions During Melting of the MSRE Fuel Salt, ORNUfM-13285, Lockheed Martin Energy Systems, Inc., Oak Ridge Natl. Lab., in preparat ion. WSRC-RP-94-396 N.d. Nuclear Criticality Safety Analysis Summary Report: The S-Area Defense Waste Processing Facility.
ORNL/ER-380 DISTRIBUTION
- 1. J.C. Anderson 25. D.M.Matteo
- 2. L. V. Asplund 26. L. E. McNeese
- 3. C. E. Bamberger 27. G. T.Mei
- 4. E.C.Beahm 28. T. C. Morelock
- 5. M. E. Buchanan 29. D. E. Mueller
- 6. S.N.Burman 30. P. T.Owen
- 7. S. G. Coffey 31. B. D. Patton
- 8. A.G. Croff 32-66. F. J. Peretz
- 9. E.C.Crume 67. L.P.Pugh
- 10. . G. D. Del Cul 68. D. W.Ramey
- 11. J.R.DeVore 69. D. G.Renfro
- 12. J. R. Engel 70. J. B. Richard
- 13. R. L. Faullmer 71. D. J. Richards
- 14. C. W. Forsberg 72. J. E. Rushton
- 15. M.A.Green 73. P.R. Sanders
- 16. E.M.Hodges 74. P.A. Schrandt
- 17. F.J.Homan 75. R.M. Szozda
- 18. J. S. Ivey 76. P. B. Thompson
- 19. P. S. Johnson 77. L.M. Toth
- 20. B.L.Kimmel 78. J. R. Trabalka
- 21. R. A. Kite 79. K.L. Walker
- 22. M. W. Kohring 80. D. F. Williams
- 23. C. T.Kring 81. Central Research Library
- 24. J. R. Lawson 82. ER Document Managemnt Center-RC
- 83. J. J. Laidler, Director, Chemical Technology Division, Argonne National Laboratory, Building 205, 9700 South Cass Avenue, Argonne, IL 60439-4837
- 84. P.A. Ostby, Idaho National Engineering, Mail Stop 3210, Lockheed Martin Idaho Technologies, P. 0. Box 1625, Idaho Falls, ID 83415-3210
- 85. R. D. Denney, Idaho National Engineering Laboratory, Mail Stop 3114, Lockheed Martin Idaho Technologies, P. 0. Box 1625, Idaho Falls, ID 83415-3114
- 86. J. C. Browne, Program Director, LANSCE and Energy Research Programs, Los Alamos National Laboratory, Mail Stop H845, Los Alamos, NM 87545
- 87. Tom Gutmann, Building 704-S, Room 41, Savannah River Operations Offfice, P.O. Box A, Aiken, SC 29801 88-90. David Pope, Jacobs Engineering Group Inc., 125 Broadway Avenue, Oak Ridge, 1N 37830
- 91. L. Youngblood, Advanced Integrated Management Services, Inc., 575 Oak Ridge Turnpike, Suite B-3, Oak Ridge, 1N 37830
- 92. P. Haas, Advanced Integrated Management Services, Inc., 575 Oak Ridge Turnpike, Suite B-3, Oak Ridge, 1N 37830 93-94. Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, 1N 37831}}