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 Retrieval and Pretreatment Retrieval of waste consists of removing the bulk of the waste from the underground tanks or bins and safely transporting it to the process facilities. Waste retrieval is treated in this chapter as part of the pre- treatment process and will also be discussed in Chapter 6 with regard to tank closure issues. Pretreatment consists of separating the HLW components from the bulk of the waste and preparing the waste stream for feed i ng to i mmobi I ization processes. Th is preparation is necessary to remove constituents that interfere with the HLW immobilization process and achieve the final objective of reducing the volume of solidified HLW produced and meeting quality criteria. Since most defense HLW was neutralized (with the exception of INEEL waste) to minimize tank corrosion, it contains large amounts of a wide variety of non-radioactive salts and metal hydroxides. Sidebar 4.1 describes the pretreatment programs for the HLW sites. The pretreat- ment of defense HLWis a much more complicated problem than the pretreatment of relatively pure acidic HLW from processing commercial reactor fuel (as performed in some foreign countries such as France, United Kingdom, and Japan) because HLWis not neutralized. The pre- treatment required depends on the chemical and radiochemical compo- sition of the waste. Because of the complex and varied chemistry of the elements present in the mixture, it is usually not possible to find a sin- gle reagent or chemical process efficient enough to accomplish all of the separations needed. Therefore, multistep processes are normally required. The unit operations that are performed as part of pretreatment (Figure 4.1 ~ include the fol lowing: · retrieval and blending; solids-liquid separation; sludge leaching; and · liquid decontamination. R e t r i e v a I a n d P r e t r e a t m e 37
SIDEBAR 4.1 SITE-BY-SITE PRETREATMENT BASELINES Pretreatment at the Hanford Site Retrieval of waste at the Hanford Site has just begun; in fact, to date, none of the tanks has been com- pletely emptied. After retrieval, the alkaline waste will be sent to a pretreatment facility to separate hazardous constituents from bulk material.The SLS process is followed by enhanced sludge washing. Cesium and strontium will be removed by ion exchange from the liquids and sent to the HLW immobi- lization facility along with the HLW solids.The Hanford Site has selected vitrification in borosilicate glass for immobilization of the LLW stream, as well as the HLW. Pretreatment at INEEL The proposed plan includes the "non-separations alternative early vitrification option,"which encom- passes (1 ) direct vitrification for the liquid SEW, with use of the current tank farm ending by the end of 2012, and (2) for the calcite, enhancement of characterization, retrieval, and treatment technology, leading to vitrification with or without separations by 2035.The method for processing and disposal for calcite will depend on the results of future separations research (Huntoon, 2000). A previous NRC committee on alternatives for HLW treatment at INEEL concluded that calcite is chemically stable and safely stored; therefore no immediate action should be taken (NRC, 1 999b). Pretreatment at SRS The alkaline salt solutions retrieved from different tanks are mixed in pretreatment tanks to"blend down"components that might cause problems in the vitrification step.The sludge is separated and pretreated by caustic leaching to solubilize aluminum from the HLW feed.The SRS is still in the process of selecting a method to separate cesium from the supernate, dissolved saltcake and sludge-leach solutions.The previous method (simultaneous large-scale in-tank precipitation with tetraphenylborate to remove cesium, and sorption with hydrous sodium titanate to remove strontium) was discarded because of technical difficulties related to rapid release of benzene (NRC, 2000b). Studies to resolve the benzene problem, as well as to identify alternative processes, are currently under way. In the mean- time, HLW sludge is being vitrified for future disposal (see Chapter 5). Pretreatment at SRS also includes adding formic acid to reduce mercuric compounds in the waste to metallic mercury.This mer- cury is then steam-stripped from the feed directed to the immobilization facility to avoid its presence in the melter. Pretreatment at WVDP After the sludge remaining at the bottom of the tanks was washed, the wash supernate and the origi- nal supernate liquid were pretreated with zeolite to remove the radioactive cesium, strontium, and residual transuranic elements.The decontaminated supernate was concentrated by evaporation and immobilized with cement.The high-level sludge and zeolite were combined and vitrified in borosilicate glass.The WVDP has retrieved and vitrified more than 95 percent of the initial waste. Zeolite was not selected for treating supernates at other sites because it would generate an excessive volume of vitri- fied HLW. ·~e H ~ G H - L E V E E W A S T E 38
Water · Retrieval Process Slurry ~ ~ ~ Liquid 1K ~ ~ ~ HEW Storage Tank HEW Storage Tank Retrieved Slurry Slurry Feed Storage Tank Slurry Solids-Liqujd | Slurry Separation I ·| Leaching Pretreatment Processes Solution Blending Tank Washed Slurry Solids Ljquid I Solids · Separation Solution I A_ Separation Processes Cesium Strontium Transuranic elements Technetium Other Elements I LLW l l HL~ I m mobi I ization I m mobi I ization + Disposal Disposal The first step consists of the retrieval of waste slurry from the tanks and the mixing of various batches of feed materials to achieve a reason- able volume of feed with uniform properties. SLS is performed to sepa- process. rate solids from the slurry, leaving a solids-free solution. This separation may recur at several points in the process. Sludge leaching consists of R e t r i e v a I a n d P r e t r e a t m e FIGURE 4. 1 Simplified scheme of the pretreatment 39
the selective dissolution of specific constituents from the solids in the retrieved waste slurry. Liquid decontamination consists of the selective removal of radioactive and hazardous species from the liquid streams. The solids from sludge leaching and the species removed during the liq- uid decontamination step are combined to generate the feed for HEW immobilization. If key radionuclides have been adequately extracted, this waste is incidental to reprocessing, or WIR, and it is classified according to its composition (see p. 1 1~. In addition to small concentra- tions of radioactive species, WIR may contain significant amounts of hazardous wastes (such as chromium, mercury, and lead) as well as large amounts of soluble salts (such as NaNO3, NaOH, and NaAIO2~. Additional constraints are imposed if WIR contains hazardous material causing it to be classified as mixed waste, as defined in Sidebar 2.2. Disposal of WI R. although less expensive than that of H LW, sti l l repre- sents a high cost because of its large volume. Pretrealment Issues Pretreatment operational issues, long-term research needs, as well as general approaches used at the different sites are described below. The objective of the long-term basic research recommendations for pre- treatment is to provide the scientific basis for developing high-efficien- cy, high-throughput separation methods that could reduce HEW pro- gram costs over the next several decades. Retrieval and Blending To date, waste has been completely retrieved from two tanks, both located at the SRS. The DOE plans to retrieve the remaining waste from the storage tanks with methods used in the past at the Hanford Site and SRS. Most of the waste will be retrieved from the tanks by pumping out the slurry of supernate and solids. In the case of INEEL waste, calcine will likely be retrieved from the storage bins by vacuuming. Stabilized water (water adjusted for pH and REDOX potential to minimize corro- sion of the steel tank) will be added to the residual waste. Following mixing to Dissolve soluble components e.g., saltcake containing alkali metal salts including cesium) and suspend insoluble solids (e.g., hydroxides of iron, chromium, nickel, transuranic elements, alkaline earth sulfates, and phosphates) the slurry will be pumped out as before. Recovery of residual waste when the tank is nearly empty will be done by sluicing, which may present problems in the case of leaking tanks, depending on the extent of recovery required. . . . .. . . . . . . . . . . . H ~ G H - L E V E E W A S T E
Wastes from multiple tanks can be blended in a waste receipt tank, as currently done at the SRS, but this may not be done at the Hanford Site because of the lack of free tank space. Instead, the Hanford waste will be treated on a tank-by-tank basis (see Chapter 5~. Blending of the waste feeds can be used to reduce the final volume of immobilized waste, by diluting those components that (1 ) have low solubility in borosilicate glass, which limits the achievable waste loading, and (2) have elevated concentrations in some tanks. The "blended" compo- nents will therefore be less limiting with regard to waste loading. In general, the waste will reside in various tanks for an extended period prior to transfer for pretreatment, which will allow time for mixing, sampling, and analysis to ensure that it meets the feed specifications for the pretreatment steps, discussed below. tong-Term Research Need After some discussion, the committee decided that blending and retrieval is not a fertile topic for basic research because it would over- lap with other research activities undertaken within the EM, such as those of the Tanks Focus Area (TEA) (DOE-TFA, 2000b). So~ids-[iquid Separation Solids-liquid separation is required to physically separate the insolu- ble components from the supernate and leach solutions. This is a key operation because most of the hazardous material (e.g., strontium, transuranic elements, but not cesium) is associated with the solids. Since the clarified liquid must meet the stringent decontamination requirements for LLW, it is necessary to remove solids to an unusually high extent. For example, decontamination factors (DFs) in the order of 1 0,000 to 1 00,000 may be requ i red for some red ionucl ides to meet specifications for the immobilized LLW product; but solids removal by such large factors is not commonly achieved in practice in a single process cycle. Solids-liquid separations will likely be required at more than one point during the process. Although simple settling and Recantation will be adequate for some processes, it is necessary to remove the solids with an extremely high efficiency, at least once, during the process cycle to meet the decontamination requirements for all of the con- stituents i n the clarified I iqu id. I nadequate SLS wi 11 i m pact downstream operations for example bv causing olu~in~ of ion-exchange columns , , , , , in, w~ ~ Ratio of concentration of a species in the feed to that in the effluent of a process (see the glossary in Appendix G). R e t r i e v a I a n d P r e t r e a t m e
and by degrading DFs for most processes. This operation is critical to the operability and performance of pretreatment, and its success is not assured. tong-Term Research Needs Long-term basic research is needed on filter media material and on filtration methods using appropriate simulated slurries followed by veri- fication using actual waste. The objective is to improve the rate of filtra- tion, filter media life, resistance to plugging or fouling, and removal efficiencies for very small solid particles and colloids (removal factors at least 104 with a range of particle sizes down to 0.1 micrometer, and possibly colloidal). Additionally, SLS methods other than filtration (such as centrifugal separation or flocculation and settling) should be investi- gated. Sludge [caching After being separated from the retrieved slurry, the sludges are leached by mixing them with a solution of sodium hydroxide to dis- solve some of the bulk constituents. The purpose of leaching is to remove compounds that would be detrimental to the immobilization process or to the quality of the immobilized product. Examples of such compounds are sodium and aluminum, which are large contributors to the bulk waste material, and chromium, sulfate, and phosphate, which, although present in much smaller amounts, are relatively insoluble in borosilicate glass and interfere with vitrification operations. If not removed, these non-radioactive materials would increase the final vol- ume of the HEW and/or lead to a hard-to-process slurry feed or an unacceptable final product. Removal of these undesirable species from the sludges is currently based on the "enhanced sludge leaching" method that is being used at the SRS and will likely be applied at the Hanford Site. This technique consists of leach i ng sl udges with strongly al kal i ne sol utions with the purposes of (1 ) solubilize aluminum and some other elements and (2) metathesize2 and thereby partially solubilize anions such as phosphate and sulfate from insoluble salts (e.g., Cased. The main problem with the enhanced sludge leaching method is that a large fraction of the waste (hydroxides of iron, manganese, nickel, zirconium, and other metals) cannot be dissolved in alkaline conditions. An alternative approach is to leach with acidic solutions to dissolve more of the solids, thereby leaving very little sludge. This was considered for Metathesis is a double-decomposition chemical reaction of the type AB + CD ~ AD + CB that is driven by the law of mass action when there is an excess of one ~on. H ~ G H - L E V E E W A S T E
Hanford waste (Swanson, 1993) and is still under consideration for INEEL waste, because alkaline leaching is not effective for the zirconi- um-type calcine (NRC, 1 999b). In fact, except for limited cases involv- ing complexing with concentrated fluoride or oxalate, zirconium can be maintained in solution only under acidic conditions. Zirconium is the dominant element that has to be separated only for INEEL waste; however, in all cases, the leach solutions require a subsequent SLS step. Operations using precipitation and solids leaching can be affected negatively by several factors: · the presence of high salt concentrations and organic complex- ants that may solubilize strontium, and perhaps transuranic ele- ments; · the solubility and speciation of actinides (plutonium, neptunium, americium, uranium) and strontium in caustic media containing carbonates and organics (e.g., oxalic acid); and · secondary reactions in leach solutions that can lead to inadver- tent reprecipitation and gel formation. These reactions have been observed in some aged or mixed Hanford leach solutions that were saturated in many components. A number of unknown reactions could occur during solids leaching and washing, with uncertain consequences for the chemical and physi- cal properties of the waste stream. Slow reactions can occur between aluminates, silicates, and other materials to form complex solids. Leach solutions and clarified supernates may be mixed together and/or evapo- rated, thereby changing their composition. Since leach solutions were initially saturated in several components, it is not surprising that new insoluble compounds can be generated. This has two main implications for the processing: (1 ) SLS processes must be performed several times, and (2) solids can continue to precipitate, which can disrupt operations such as filtration, ion exchange, and evaporation. A large amount of research has been done on solubilities in multi- component systems, for example by RapLo and Lumetta (2000), and efforts have been made to model the results for application to the extremely complex waste systems. At Hanford, for example, a thermo- dynamic equilibrium program called Environmental Simulation Programs is used for several purposes, such as estimating the conse- quences of waste mixing and dilution during transfers and retrieval (Mahoney et al., 2000; Papp, 1998~. Extension of such models to include kinetic data and solid phase identification would be beneficial. At INEEL, there are only very limited data on dissolution of calcine in acid solutions, and more studies are needed if the waste is to be processed before i mmobi I ization (N RC, 1 999b). R e t r i e v a I a n d P r e t r e a t m e n t
tong-Term Research Needs Sludge leaching offers several long-term basic research opportunities for successful operation of the pretreatment processes for all of the sites. Projects should focus on chemical kinetic data (rates of dissolu- tion) and equilibrium data (residual concentrations and solubility prod- ucts) in very complex multicomponent systems. Examples of such com- ponents are: sodium, aluminum, silicon, iron, chromium, zirconium, sulfate, phosphate, carbonate, hydroxide, nitrite, and nitrate, many of them dissolved at saturation limits. In spite of the complexity of the task, research in this direction can yield useful results in the long term. Research is also needed on solution stability and precipitation mecha- nisms in highly concentrated solutions. Of particular interest is the sta- bility of solutions with time, relative to reprecipitation of new phases and gel formation. To the extent practical, effects of radiolysis should be i n c I uded i n expert ment with s i m u I ants. Long-term basic research is needed to develop methods to enhance removal or to cope with the presence of problematic material and con- stituents not readily dissolved by alkaline leaching (e.g., decomposition of organic materials and dissolution of Cr3+ by oxidation to Crop. In addition, the option of acid leaching of the sludge should be investigat- ed to dissolve a greater fraction of the solids. Finally, research leading to the development of a predictive model for solids-liquid behavior in concentrated salt solutions could result in reduced costs and time. Liquid Decontamination After sludge leaching and SLS, the supernate and all leach solutions are carried through one or more processes designed to remove specific hazardous or radioactive constituents so that the liquid product can meet the acceptance criteriafor LLW immobilization and disposal. As indicated in Figure 4.1, these processes may include removal of radioactive materials such as cesium, strontium, transuranic elements, technetium, and other elements, as well as non-radioactive toxic metals and organics. Several processes are known for each of these separa- tions, such as zeolite ion exchange to remove cesium and strontium and hydrous sodium titanate ion exchange to remove strontium and transuranic elements. Since none of the processes is as simple and effective as might be desired,3 there is continuing interest in finding improved separation methods. 3For instance, the monosodium titanate process employed at the SRS may not be successful for all waste streams since it may require extensive blending because of low DFs. H ~ G H - L E V E E W A S T E 44
There is potential advantage in combining two or more separation processes into a single unit operation, for example, by combining processes for both strontium and transuranic elements removal, as well as for finding a single reagent that effectively will remove more than one problem constituent. An example of a combined process is the use of a mixture of chlorinated cobalt dicarbollide, polyethylene glycol and octyltphenyl)-N, N-d i isobutylcarbamoyl methyl phosph i ne oxide (CMPO), which has been used in Russia for simultaneous removal of cesium, strontium, and transuranic elements and has received limited testing at INEEL (Thompson, 1998~. Absorption few., using activated carbon), coprecipitation, and ion . . . ~ ~ . exchange provide selective removal of contaminants, either on particles in columns or on absorbent solids that are added to the solution and then removed by SLS. The main issues to be addressed with absorption methods are the maximum achievable decontamination effectiveness and the amount of sorbent required to achieve such effectiveness. The loaded sorbent must either be routed to the HLW for disposal or be eluted and reused for subsequent loading cycles. If reused, the attain- ment of sufficiently large DFs (in the range up to 105) in subsequent cycles is usually very difficult. Solvent extraction is another option: a solvent extraction process using a crown ether is under study for SRS to remove cesium from the HLW liquid (NRC, 2000b). Solvent extraction may be particularly advantageous for acidic solu- tions. Because of the unique composition of the INEEL HLW, the caus- tic leaching process used at the other sites offers little benefit. Acid leaching and separation of inactive components from the acidic solu- tion is an option. The final immobilized HLW volume could be decreased substantially if acid leaching and acid-side processing (Swanson, 1 993) were selected for either Hanford Phase 11 or INEEL cal- cine. The value of this approach depends on the availability of reposito- ry space for DOE HLW. Current plans call for HLW from Hanford and SRS to go to a first geological repository. HLW from INEEL will be directed to a second repository, particularly if processing is not com- pleted before the closure of the first repository (NRC, 1 999b, page 83~. However, at present the United States is investigating only one site (Yucca Mountain) to determine its suitability to host a geological reposi tory, and a decision to proceed has not yet been made. The possibility of a second repository is h igh Iy conjectu ral at th is ti me. In all cases, each separation process generates some quantity of sec- ondary waste, such as ion-exchange sorbents, organic solvents, or vari- ous conditioning solutions. Dealing with these secondary waste streams is sometimes difficult because they can interfere with the separation processes and increase final immobilized waste volume. Therefore, it is necessary to develop processes that minimize secondary wastes or gen- R e t r i e v a I a n d P r e t r e a t m e n t
erate secondary wastes that are relatively innocuous. The strategies for pretreatment end for both HLWand LLW immobilization and disposal are mutually interdependent. Because of this connection, the nature of both immobilized wastes will be determined by the separation and pre- treatment strategy adopted. An important research effort in the pretreatment area is already car- ried out within the Efficient Separations Program-lntegrated Program (ESPI P), a crosscutti ng project with i n the EM (DOE-ESPI P. 2001 ). Its mission is to identify, develop, and perfect separation technologies to separate cesium, strontium, and transuranic elements from radioactive waste streams. Additionally, the EPA has developed a large database of technologies that might be applicable for the removal of non-radioac- tive toxic materials (metals and organics from the HEW and LLW). The EPA Superfund Innovative Technology Evaluation (SITE) program evalu- ates all available information on the technology for hazardous waste remediation and analyzes its overall applicability to other site charac- teristics, waste types, and waste matrices. The objective of the SITE pro- gram is to encourage the development and implementation of (1 ) inno- vative treatment tech nologies and (2) man itori ng and measu remeet. For further information see EPA-SITE (2001~. tong-Term Research Needs A long-term basic research effort is needed to complement the activ- ities within the ESPIP to identify sorbents and separation methods for cesium, strontium, technetium, and transuranic elements that are effec- tive and operable in high-ionic-strength alkaline or acidic solutions and high radiation fields. Sorbents must have high selectivity and capacity and must either be capable of elusion and regeneration for a number of cycles or be easily decomposed, such that disposal of the sorbent has a minimal impact on the final volume of immobilized waste. For exam- ple, inorganic sorbents should be investigated if their selectivity is suffi- ciently h igh (e.g., monosodium titanate for strontium and acti n ides, ammonium molybdato-phosphate for cesium). Decontamination methods that combine removal of more than one target constituent within a single unit operation (e.g., a combination of sorbents or extractants that can remove cesium, strontium, and transuranic elements simultaneously from a feed composition) are needed and warrant further basic research. Research in solvent extrac- tion separation methods is needed to identify stable, selective, high- capacity extractants that also are inexpensive and commercially avail- able. In addition, research on solvent extraction methods would be par- ticularly valuable for use with acidic waste solutions for which solid sorbents are often less effective. Solvent extraction from acidic solutions is the standard for reactor fuel reprocessing, because fuel is soluble H ~ G H - L E V E E W A S T E 46
only in acidic solutions. The waste treatment methods now being used are generally alkaline, mainly because the feed is usually alkaline and there is reluctance to add large quantities of nitric acid to re-acidify it. However, investigators should not restrict consideration to alkaline methods only, even though that is the starting medium. There could be benefit from treating solutions from acid leaching of alkaline sludge or INEEL calcine, but the methods would be very different from fuel repro- cessing. R e t r i e v a I a n d P r e t r e a t m e 47
