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.
4 Preferred Technical Approach Based on present understanding of the condition of the system and its salts, a number of observations are offered that are relevant to any decision on remediation options and strategy. As further information is obtained, these observations may have to be modified. These comments form the basis for a preferred technical approach developed by the panel. COMMENTS ON PROCESS STEPS As discussed earlier, it is possible to conclude the following: 1. Since the charcoal trap is "at the end of the line" (Peretz, 1996c, Figure I.9, p. I-13) and all the migrated uranium salts originated in the drain tank, it is prudent to assume that all the piping contains uranium in volatile and nonvolatile forms such as uranium hexafluoride (UFO; solid and gas), urany! difluoride (UO2F2), uranium pentafluoride (UFs), uranium tetrafluoride (UF4), and uranium trifluoride (USA. 2. Since the freeboard gas space above the salt is at a lower temperature than the salt, there are probably significant uranium compounds deposited on the walls of the tanks and the thimbles and on the underside of the tank cover. 3. Current plans to resolve issues 1 and 2 above (with collection of uranium compounds in the piping and the auxiliary charcoal bed tACB], for use in a mass balance estimate) are important to determine the 1The panel is aware that Oak Ridge National Laboratory (ORNL) documents (e.g., ORNL, 1996b) have used various assumptions about technology and process steps for remediation. It is the panel's understanding that these assumptions were made for administrative reasons such as budgeting and staffing and do not represent actual decisions for implementation. 45
46 AN EVALUATION OF DOE ALTERNATIVES FOR MSRE appropriate strategy for dealing with the uranium in the salts. Indeed, it is conceivable that there may not be as much uranium left in the salt as initially anticipated; the uranium may be concentrated in certain regions exterior to the bulk salt. A directional, high-resolution gamma survey (e.g., using a welI-shielded, small-bore collimated germanium [Ge] detector) could help clarify this question. 4. Separation of uranium from the salt and from the system should be one of the major milestones, because it would eliminate subsequent criticality concerns and reduce the rate of radiolysis. Doing this without first moving the salt mixture to another location is attractive because it avoids the hazards (and costs) of any transfer operation and avoids leaving a residue of residual uranium and salt that is inherent in such liquid salt transfers. 5. Except for fluorination in place, all other separation technologies appear to require extensive new facilities with resulting long times and high costs. With the present state of knowledge, the pane} considers that some type of fluorination procedure to extract UFO from the salt is the most promising technical approach to address comment 4 above. DEVELOPMENT OF A PREFERRED APPROACH With consideration given to the present state of knowledge of the condition of the Molten Salt Reactor Experiment (MSRE) system and salts and to the above-mentioned list of technical approaches that seem credible, several information-gathering exercises are recommended to assist the decision-making process, as discussed further in Chapter 8. Subject to confirming evidence yet to be gathered, the panel recommends that the following considerations, outlining a cautious, stepwise strategy, be taken into account in the development of a preferred approach: 1. Additional chemical, physical, and process data on the distribution of uranium and the remelting and refluorinating behavior of the salts, as discussed in Chapters 2 and 3, are essential to validate assumptions and reduce uncertainties; therefore, the plan must be developed in stages either as the information becomes available or as it becomes clear that such information cannot be made available.
PREFERRED TECHNICAL APPROACH 47 2. Venting of the off-gas system piping and the gas space above the salt in the drain tanks, together with volatile UFO removal from these areas, is an essential first step for any alternative, and the resulting reduction in system pressure will be a significant safety step. The collection of uranium will also provide the first data point in an ongoing material balance program. 3. The next step depends on the success of step 2 and an assessment of system integrity. If possible, it should involve removing the presently nonvolatile uranium salts from the piping and tank freeboards and obtaining a material balance. 4. During steps 2 and 3 above, noninvasive methods of locating significant uranium salt inventory (e.g., high-resolution gamma spectroscopy scans) should be undertaken as feasible. 5. After "cleaning" of the piping and tank gas space, obtaining a sample of the fuel salt could be considered. This sample seems useful for evaluating the methods and possible hazards of salt melting, but such merits should be weighed against the hazards of taking a sample and of contamination problems. Specifically, the gaseous radon hazard (see Appendix C), and contamination with alpha- and beta-gamma-emitting residues, should be considered. With an inventory of 5 g of uranium-232 (232U; 130 curies) in the salt, a I-pound sample (0.01 percent by weighty would contain approximately 13 millicuries (mCi), not only of224 U. but al2s20o of the decay products thorium-228 (2 Pith), radon-224 ~ Rn), and Rn, the latter of which is a highly radioactive, alpha-emitting gas. The 220Rn gas permeates neoprene gloves and distributes rapidly to the air even in countercurrent air flows. It then decays rapidly to polonium-216 (typo) and on to a 3.3-hour half-life daughter, lead-212 `2~2Pb), that finally decays (with a 36 percent yield) to thallium-208 ~ TI), which emits a 2.6-MeV (million electron volt) gamma ray (see Figure 2.~. 6. The alternatives for salt melting include the following. but choices cannot be made until more information is available: a. Melt as is. melt. A, b. Hydrofluorinate below the liquidus temperature and then
48 ANEVALUATION OF DOE ALTERNATIVES FOR MSRE c. Use pool melting by either heater or laser, and hycirofluorinate as melting occurs. d. Perform step 6b without melting, and then step 6c. 7. At this point, if melting seems feasible, the decision can be made whether to fluorinate in place or to transfer the molten salt to another vessel for fluorination. a. If the decision is to fluorinate in place, choices must be made concerning fluoridating agents (see Chapter 3 and Appendix B), melting methods (all at once or stepwise), and whether to use existing =7 ~ ~ ~ tank heaters or a new heat source. b. If the decision is to fluorinate elsewhere, the choice must be made as to where and how. Hydrofluorination in place, followed by melting (to test the salt solubility), followed by transfer of the melt to a new vessel of ensured integrity for fluorination is a strategy with merit if corrosion concerns persist, because the hydrofluorination step, which may enhance the melt, is a less corrosion-aggressive process than fluorination. The options of thoroughly inspecting the vessel walls for localized thinning, and possibly plugging the thimbles or filling them with HF-resistant material, could be used to reduce any likelihood of leakage. 8. If melting either does not appear to be possible or does not appear to be safe, the solid removal alternative would have to be developed. In the development of steps 6 through 8, some laboratory and mockup tests and use of these test results in deciding what large-scale process to apply to the drain tank salts would seem to be in order. For example, the flush salts might be a preliminary surrogate, but because of the (assumed) large difference in constituent concentration and oxidation state, either melting or fluorination of the flush salts is only a demonstration of the equipment and technique and does not confirm the chemistry of the fuel salts. If all else fails and solid salt removal becomes the proposed operation, it may be necessary to conduct a mockup test to determine the mechanical problems (such as those introduced by the cooling thimbles; the mechanical challenge is greater to remove solids
PREFERRED TECHNICAL APPROACH 49 from regions interior to the tank that are behind cooing tube arrays, as compared with regions directly under a flange on the tank head). Because of the uncertainties inherent in a solid removal operation, it would be prudent to do the work in stages and not to commit to doing the drain tanks lentil after the mockup and the flush tanks have been done successfully. A particularly significant issue in carbon dioxide (CO2) blasting or other solid removal techniques is gaseous radon contamination. Whereas fine particles of salt can be filtered from a gas stream, radon cannot. This consideration alone could well preclude the use of such means of breaking up the salt if fluorination (with all the caveats regarding melting) fails. Also there should be efforts to determine the status of the system (leak-tightness, pressure worthiness, corrosion damage to date) and projected damage, as well as complete assurance of the absence of water from the cells or the systems (since essentially all the credible postulates of criticality excursions require the presence of water). Tests on surrogate samples can be useful to address some of the uncertainties associated with the fuel salts. For example, such tests could be used to bound the range of redox conditions finder which hydrofluorination and fluorination would be successful and to probe the extent of segregation that could be tolerated. Peretz (1996c, p. I-30) noted that segregation was observed on freezing surrogate salt samples (uranium concentration varied by a factor of two, from top to bottom). This effect would likely have been enhanced by the slow (melt refining) process of cooling the several tons of irradiated salt. Radiation effects and transport of UFO from the mass may have further increased the inhomogeneity of the uranium and fission product distribution. A sample of the actual fuel salt, preferably a core sample down through the mass, would make it possible to test several remedies on the actual salt itself before committing to major action that could be irreversible. However, as noted earlier, the decision to obtain a core sample should consider the disadvantageous consequences of possible 22sTh 224Ra and 220Rn contamination Chapter 5 discusses specific technical processes mentioned above that are finder consideration (Peretz, 1996c) for stripping uranium from the salt.