Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
9 Summary and Responses to Questions in Statement of Task The Statement of Task for this study (Box ~ . T) asks the pane! "to review the current state of the Department's (DOE's) evaluations of alternatives" (presented in Peretz, 1996c) "for removal, separation, and stabilization of MSRE salts to determine the extent to which (~) appropriate technologies and options have been identified and evaluated; (2) evaluations are sufficiently complete to form a basis for decision- making; and (3) potential hazards associated with fuel and flush salt removal have been identified and addressed." Following a technical summary and some general perspectives that summarize pane! views, specific answers to these questions are presented. TECHNICAL SUMMARY In brief, the Molten Salt Reactor Experiment (MSRE) salts contain uranium, transuranium elements, and fission product radioactivity, as well as fluorides of lithium, beryllium, and zirconium, and should not remain indefinitely in the drain tank cell where they have been kept for more than 25 years. Major reasons for this conclusion include the continuing migration of fluorine and uranium from the bulk salt, its location below the natural water table, and the potential critical configuration upon the intrusion of water. The permanent disposition of transuranics well in excess of 100 nanocuries per gram at shallow depths, just below grade, is also likely to be disallowed. Removal alternatives for the salt are 1. melting, fluorination to form uranium hexafluoride (UF6), and removal of the molten salt with its plutonium and fission products; or 79
80 particles. AN EVALUATION OF DOE ALTERNATIVES FOR MSRE 2. breaking up the bulk fee! salt and removing it as solid The first alternative is preferred, with or without preliminary uranium removal in situ, with option 2 as a backup procedure. Radiation effects continue to move fluorine and uranium from the bulk salt. In October of 1996, there was nearly one atmosphere of excess gas pressure over the fuel drain tanks, almost twice that observed in 1994. The pressure is contributed largely by molecular fluorine (F2) and the saturation pressure of UFO which suggests solid deposits in the piping of UFO and its radiation-induced lower fluorides. Any removal alternative must start with the gases, including whatever uranium can be removed by pumping. The uranium material balance is poorly known, especially the amounts in and the locations of nonvolatile forms. Partial decomposition of solid UFO by the effects of alpha radiation (from 233U and daughters) is expected to form nonvolatile uranium fluoride deposits essentially throughout the equipment. Because fluorine is more effective in reforming UFO when such locations can be heated, alternate reagents for UFO formation (e.g., BrFs [bromine pentafluoride] and KrF2 Krypton difluoridel; see Appendix B) should be considered to treat nonvolatile plugs in unheated regions. Removal of uranium from the molten Ale} salt by F2 sparging to produce UFO was accomplished routinely in the past. Hence, this approach would normally be favored. However, melting of an irradiated] surrogate salt sample did not yield a clear melt but yielded at least two phases, one a metallic-appearing precipitate. To date, it has yet to be established that refluorination treatment of the salt containing such precipitates will be successful in reestablishing a homogeneous melt. Experiments should be performed on additional irradiated surrogate salt samples in attempts to redissolve the precipitate and restore the salt melt to a homogeneous condition so that transfer to another container or uranium removal by fluorination could ensue. It should be emphasized that the MSRE fuel salt, probably in a highly reducing chemical condition, remains a major problem, which will only worsen with time. Even if uranium removal is accomplished, the residual salt will still contain activity from 0.7 kg of plutonium and from fission products. Formation of fluorine and volatile fluorides will continue. Plans to
SUMMARY 81 "getter" the fluorine formed from the interim stored fuel salt should also provide for the possible formation of some plutonium hexafluoride (PuF6), although a significant quantity of volatile PuF6 is not expected (Mills, 1996~. STRATEGY FOR REMEDIATION There is insufficient knowledge in hand to outline, in detail, a series of processes that, if followed, would lead clearly to a satisfactory outcome of nonhazardous removal, separation, and stabilization of MSRE salts. Information gathering in the near future should allow better decisions concerning these alternative actions. Indeed, relevant alternatives have been identified in Peretz (1996c), but final evaluation awaits further information. No important options or technologies seem to have been omitted from Peretz (1996c). There is a preferred approach within which some options exist. Of the several alternative processes, fluorination to extract UFO from the molten salt is the leading option. However, choices are available for specifics of the processes for salt pretreatment, salt melting, and fluorination. Although Peretz (1996c) contains a baseline hazard section that identifies present hazards, evaluations are not yet sufficiently complete to allow for thorough knowledge of the extent of hazards associated with various alternative processes. However, the information required is well defined and can be obtained by available methods. Three information needs are (1) locating the uranium within the system, (2) further work in assessing the condition of the fuel salts and their containment, (storage tanks, piping, and associated hardware), and (3) bench-scale testing of possible remediation processes. Where Is the Uranium? As the charcoal trap and the piping system are cleaned up, some related information will be obtained. As the volatile components are taken from the system, identified, and quantified, more information will follow. The amount of uranium removed from the piping as the hexafluoride plus that in the charcoal trap will help determine the
82 ANEVALUATION OF DOE ALTERNATIVES FOR MSRE remaining total inventory. This inventory could be present in lower oxidation states, either in the salt or deposited in the headspace and piping. If there is much uranium remaining in the system external to the tanks, it might be feasible to map its distribution by radiation surveys. The distribution of uranium-containing species in the salt could be determined with gamma monitoring supplemented by neutron monitoring. Of possible relevance to a remediation strategy is whether inhomogeneities in uranium concentration exceed the variations in distribution that were observed in surrogate salt samples. Core and/or radial samples could also provide this information, but gaseous radioactive contamination is a significant hazard associated with sampling the salt. How Can a Condition Assessment Affect Remediation Plans? Information related to the condition of the system, obtained as work has progressed (ORAL, 1996a), has established that plugs from deposits in the piping remain and interfere with the removal of volatile substances by pumping. Fluorination may remove lower oxidation state uranium that could remain in the plugs and headspace. Although fluorine, hydrogen fluoride (HF), and BrF5 are the preferred fluorination agents, corrosion and temperature concerns and the extent to which uranium-containing deposits resist reaction may lead to consideration of other fluoridating agents, such as atomic fluorine or KrF2 (see Appendix B). Plutonium is present in such low concentrations (approximately 155 ppm tparts per million]) that it is not likely to form PuF6 readily with any of these possible fluorinating agents. Although it seems unlikely since gases above atmospheric pressures remain in the system, there may be cracks or pinholes in the apparatus and piping above the salt. These would not be revealed if they were plugger! by the uranium-containing species. If it were determiner! that corrosion damage is not a problem, consideration could be given to plugging the thimble tubes physically and fluorinating the gradually melting salt in the storage tanks. The panel recommends, prior to 'Fluorine is a better oxidizer at elevated temperature. If the pipes and valves cannot be heated, fluorinating agents more active than F2 at ambient temperature may be required.
SUMAl4RY 83 fluorination, consideration of an annealing procedure of heating the solid salt in an atmosphere with a gradually increasing concentration of fluorine or HF. The panel recommends an initial atmosphere with an HF- inert gas mixture, rather than the proposed HF-H2 (hydrogen) mixture. This would be succeeded by a ~-F2-inert gas mixture to ensure that all uranium was oxidized at least to uranium(IV). Omitting hydrogen from the hydrofluorination step would reduce the hazards associated with the explosive reactivity of hydrogen gas.2 The Panel's Preferred Alternative Because one does not know at this time what options will be available for final disposition of the actinides, fission and decay products, and the salt, the panel considers that interim storage of separated uranium and of the residual salts is the only realistic option available at this time. Interim storage of the separated uranium as an oxide in existing sites at Oak Ridge seems acceptable, as does interim storage of the fluoride salts (containing radioactive species other than uranium) with a getter for radiation-produced fluorine gas that would be expected to form and migrate out of the solid salt residue. in short, the pane! finds that the general approach given in Peretz (1996c) is acceptable but backup alternatives must be available. The scheme should have hold points to confirm that the expected process behavior is being realized. Given adequate preparations, the intended processes are likely to perform as planned. However, it would be prudent to add faliback and contingency plans to the intended operational scenario. Hazards The pane! considered various hazards. Criticality, in particular, was given considerable attention. The panel finds that, in the absence of water, there is no appreciable likelihood of the occurrence of a critical excursion. As stated in Chapter 6, the pane! believes that the probability 2The HF-F2-inert gas mixture would produce hydrogen via the reaction 2UF3 + 2HF ~ 2UF4 + H2. This evolved hydrogen would then be consumed by F2 to form HF. The addition of F2 thus serves to reduce the hazard of excessive hydrogen concentration.
84 AN EVALUATION OF DOE ALTE~ATIVES FOR MS~ of such an excursion during remediation of the MSRE salt is extremely low and, even if such an event were to occur, the safety and technical consequences would be insignificant. The pane! finds that the available documentation (Peretz, 1996c) shows an awareness of the hazards and how to deal with them. The panel has some concern over the use of hydrogen as a diluent and anticorrosion agent in the proposed treatment with mixtures of hydrogen and hydrogen fluoride. An evaluation of the specific hazards associated with each step of the planned work, and of the branches in the scheme to be developed, should be laid out in further detail. RESPONSES TO QUESTIONS IN STATEMENT OF TASK The Statement of Task for this study (Box A) asked the panel to address three questions concerning the scoping activities of the Department of Energy (DOE) and DOE contractors to date on the remediation techniques and strategy that could be applied to MSRE fuel and flush salts. These questions (and answers) are intended to assist the decision-making process involving DOE and regulatory agencies. Of course, selection of a proper cleanup approach is of interest to others, such as workers and the general public. The material in preceding chapters provides the necessary background for answers to these questions. The panel provides brief answers to each question below. More detailed discussions can be found elsewhere in the report. Question 1 To what extent have the appropriate technologies and options been id~entif ed and evaluated? The overall technical alternatives seem to have been identified, but the actual details and steps are not clear enough for final evaluation. in some cases, alternatives exist within a technology: for instance, if fluorination in the drain tanks is not feasible by using
SUMMARY 85 fluorine, fluorination with another agent such as BrFs or fluorination in a separate vessel may be possible. Because the source of volatile uranium fluoride was the drain tank and uranium migrated through the system to the carbon trap, there is a high probability that substantial amounts of uranium compounds are located on the cover and walls in the freeboard area of the drain tank. This issue may be important in assessing technical options. The alternative of solid salt removal by mechanical fragmentation and particle evacuation was conceptual and not specific. If there are data or experience to support the concept, they were not presented to the panel. Question 2 To what extent are the evaluations stuff cien fly complete to form a basis for decision making? The development of additional information, now in progress, is needed to support decisions about alternatives. The final selection of one method for example, direct fluorination with F2 is subject to the acquisition of additional information about the condition of the system and the hazards, which appears to be feasible given adequate resources. The literature documents the procedure, and personnel at Oak Ridge National Laboratory (ORNL) have had extensive experience in the direct fluorination of uranium tetrafluoride (UF4) to produce UFO. Nevertheless it is important to be aware of alternative possibilities. The Consequences of Failure to Complete According to Plan Additional assessment is needed to address an important hazard, the hazard offailure, to explore whether the selected remediation method precludes a desirable backup option in the event that unexpected or undesirable behavior is encountered. For instance, could the failure of one technological alternative preclude using another? Could failure to reoxidize the reduced metals sufficiently result in a sludge in the melted salt such that liquid removal becomes impractical?
86 Cost Estimates AN EVALUATION OF DOE ALTERNATIVES FOR MSRE Because of the lack of details it is not possible for the panel to assess the credibility of cost estimates. For example, it is not obvious how to compare the cost of an as-yet-undefined mechanical process with the cost of liquid removal. A Possible Strategy In ~eretz t1996c), the strategy being used by the project is not defined. Each strategy should have a primary alternative and one or more backup alternatives to cover the hazard of failure of the primary alternative. A preliminary cost estimate should be done for each case. The decision maker can then optimize the choice of strategies based on probable success, initial costs, and possible ultimate costs. The decision maker can then choose between strategies of lowest base cost but higher potential total cost versus strategies with higher base costs but lower potential overall costs. An alternative technology that is primary in one case might be the backup in another case. Question 3 To what extent have the pofenlial hazards associated with fuel andf ush sad removal been adequately iden~if ed and addressed ? The Peretz (1996c) report contains a baseline hazard section that identifies the present hazards, but there is almost no discussion of the hazards associated with various alternatives. A preliminary hazards screening has also been made (ORNE, 1995~. The term hazard is used instead of risk because the probability of a hazard becoming a risk is best determined only after operation steps are detailed for execution. Identification of the hazards associated with various process alternatives awaits further development of those alternatives. The hazards associated with radioactivity and criticality are not unique to this project. However, the complex mechanical and chemical questions elevate the importance of the hazard of failure, a concept not developed in detail in Peretz (1996c).
SUMAL4RY 87 The Peretz (1996c) report contains a cursory risk analysis of potential future human exposure in which there are indications that unduly conservative (i.e., beyond credible) assumptions may have been used. One stated assumption is a scenario in which a component of the now pressurized off-gas system suddenly fails and releases the entire radioactive inventory within the building. It is further assumed that 50 percent of this inventory is retained, and 50 percent is lost as a ground- leve! release under typical conservative atmospheric conditions (Peretz, 1996c, p. 1-36~. These assumptions are unduly pessimistic; at ambient temperature, only a small fraction of the uranium is in the gas phase, and exposure to moist air rapidly converts most of the uranium to nonvolatile oxides. In addition, all system components are in sealed cells inside a building. Any highly exothermic process with carbon (e.g., decomposition of CFX to form carbon compounds and CF4 Carbon tetrafluoride]) that could volatilize a large amount of radioactive fluoride salts is a potential hazard that has already been mitigated by isolation of the activated charcoal bed from the majority of the radioactive material. A major overestimate of a hazard is not a conservative strategy when using the results to choose a line of action. It easily can result in selecting a course of action more dangerous than the discarded action. In the absence of more complete data (e.g., realistic probability distribution functions for every uncertain parameter), and for present decisions, the panel believes that estimates of risk that provide the best basis for decision making should be on an expected value basis, bracketed by an uncertainty range. PANEL PERSPECTIVE These findings and recommendations are offered to enable the parties involved (DOE, contractors, and regulatory agencies) to clean up the MSRE fluoride salts safely and expeditiously, with due regard for the hazardous materials (such as reactive F2 and UFO gases and Missile 233U). Fluorination procedures, subject to important caveats and further information-gathering activities advocated in this report, appear at this time to be a preferred technical approach worthy of further consideration. However, without the additional information that comes from testing and
88 AN EVALUATION OF DOE ALTERNATIVES FOR MSRE experimentation, it is too early to make a final, sound decision. Thus, it is still too soon to eliminate backup remediation methods. A hazard management strategy is recommended by the pane} as an appropriate way to develop a decision-making process that is phased, or "gated." A period of further testing and information-gathering activity would provide for a more informed decision that could be deferred until the relevant information is received. Candidate technologies can be in view during this time and, subject to the receipt of confirming evidence, could be used in the MSRE salt cleanup. The aim of acquiring more information with additional scoping studies is to affect the choice of technical alternatives and enable project personnel to gain increased confidence in the success of the chosen technical approach. As new information becomes available on the fuel and flush salts and on the status of the rest of the MSRE system, additional reviews of the major issues may be warranted. OVERALL CONCLUSION After reviewing Peretz (1996c) and discussing the problems with responsible personnel, the pane} concludes that the evaluations done to date are adequate for proceeding with the remediation program and that work needs to go forward on a timely schedule. As discussed in the report, the pane} recommends a phased program with each phase designed to do those things and acquire the information that will make going on with the next phase an action with acceptable risks. The technical capabilities of Oak Ridge National Laboratory are quite adequate to ensure that the measures needed to prevent or control all likely hazards can be implemented practically and safely.